future healthcare: bioinformatics, nano-sensors, and

247

Chapter 8

Future Healthcare: Bioinformatics, Nano-Sensors, and Emerging Innovations

Shoumen Palit Austin Datta

Contents8.1 Introduction.............................................................................................2488.2 ProblemSpace..........................................................................................250

8.2.1 Background..................................................................................2508.2.2 Focus............................................................................................252

8.3 SolutionSpace..........................................................................................2538.3.1 ExistingElectronicMedicalRecordsSystems...............................2538.3.2 ChangingtheDynamicsofMedical Dataand

Information Flow...................................................................2578.3.3 DataAcquiredthroughRemoteMonitoringandWireless

Sensor Network............................................................................2658.3.4 InnovationinWirelessRemoteMonitoringandthe

Emergence ofNano-Butlers..........................................................2738.4 InnovationSpace:MolecularSemantics...................................................290

8.4.1 MolecularSemanticsIsaboutStructureRecognition...................290

K10373_C008.indd 247 10/8/2010 6:14:41 PM

248 ◾ Nanosensors: Theory and Applications

8.1 IntroductionProclaiminghealthasahumanright[1]isaplatitudewhenhealthcarestillremainsinaccessibletomillions,eveninaffluentnations,orbillionselsewhere.Fewdisputethefactthatmorethan30,000children,lessthan5yearsold,dieeveryday,manysufferingfrompreventablediseases.AccordingtotheInstituteofMedicine[2]attheNationalAcademyofSciences,healthcareissubstantiallyunderperformingonseveraldimensions: effectiveness, appropriateness, safety, cost, efficiency,preven-tion,andvalue.Increasingcomplexityinhealthcareandregulatorystepsislikelytoaccentuatecurrentproblems,unlessreformeffortsgobeyondfinancing,tofostersustainablechanges in theethos,culture,practice,anddeliveryofhealthcare. Iftheeffectivenessofhealthcareistokeeppacewiththeopportunityofprognostic,diagnostic,andtreatmentinnovation,thenthedesignmustbebasedonsystemsthinking[3].Informationtechnologymustbestructuredtoassureaccessandappli-cationofevidence,attherighttime,tofacilitatecontinuouslearning,andresearchinsights,asanaturalby-productofhealthcareprocess.Weneedtoreengineerthedevelopmentofaresearch-drivenlearninghealthcareorganization[4]byintegrat-ingasystemsengineeringapproach,tokeeptheindividualinfocus,whilecontinu-ouslyimprovingconcepts,quality,safety,knowledge,andvalue.

Inparticular,commitmenttoresearchmaybeemphasizedbylessonsfromthepast [5] to catalyze a future where creative cross-pollination of diverse conceptsfromunconventional [6] strategic thinkers is rewarded. It is equally essential tobuildmultidisciplinarycollaborativeglobalresearchteams,imbuedwiththetruespiritofdiscovery[7]inbasicandapplieddomains.Theseteamsmustbeenabledto drive an entrepreneurial [8] enterprise approach to create proof of concepts.Translationofunconventionalconceptsintorealitymaybeguidedbyinnovatorswithhorizontalglobalvisionratherthangatekeeperswhoprefertostay“inthebox”andavoidrisks that leadershipdemands.Thecaveat in thisprocess is the impa-tienceof“practical”peopleforunconventionalconcepts,butthe“nailonthecof-fin”isoftendrivenbypoliticalpolemicistswhoalsogetimpatientwithconcepts,nomatterhowjustifiedorpragmatic,becauseitgetsintheirwaytoselltheirplans[9].Thelattermayblockfundingorpolicythatmaybetterenabletheconversionofunconventionalvisionintoreality,onlylimitedbyourimagination[10].

Yet, unconventional thinkers and business leaders [11] are largely creditedfor globalization [12]. Striking transformations have occurred through decision

8.5 AuxiliarySpace........................................................................................2968.5.1 PotentialforMassiveGrowthof ServiceIndustryinHealthcare.....2968.5.2 BacktoBasicsApproachIsKeyto StimulateConvergence...........298

8.6 TemporaryConclusion:Abundanceof DataYetStarvedforKnowledge?......301Acknowledgment..............................................................................................301References.........................................................................................................302

K10373_C008.indd 248 10/8/2010 6:14:41 PM

Future Healthcare ◾ 249

systemsandprocessengineering,notonlyinestablishedmarketsbutalsoincreat-ingnewmarkets[13]despiteomnipresentglobaluncertainty[14].Thesechangeshavereshapedpoliticaleconomy[15],governments[16],theserviceindustry,andmanufacturingsector, includingbusinessprocess[17],software,hardware,bank-ing,retail,airlinesafety,automobileindustry,nationalsecurity,andthebusinessofthemilitaryindustrialcomplex[18].

Lessons from the failures and the fruits of success, enjoyed by the businessworld, may not be irrelevant for exploration and/or adaptation by healthcareorganizations,despitetheirreconcilabledifferencesthatexist inthedynamicsofmechanicalversusbiologicalandsocialsystems.Thecurrentchallengesinhealth-carecompelafreshviewoftheorganization,structure,andfunctionofthedeliveryandmonitoringprocessesinhealthcare,notonlyfortheindustrialworldthatmayaffordtheincreasingcost(macro-payments)butforglobalhealthcareservices,inamanneraccessibleto,aswellasfeasiblefor,thebillions(micro-payments).Financialsustainabilityofhealthcare is a key issue in thedesignof innovativehealth ser-vices.Thelatterevokestheparadigmofservicesthatmaybedeliverablefor“micro-payments”ratherthanthecurrentspendingthatclaimsadoubledigitshareofthegrossdomesticproductofrichnations.

Oneofmanylessonsfromthebusinessworldisthattosurviveinbusiness,busi-nessesmustexploitthepowerof“now”[19],perhapsbestillustratedbythesurgeinusingreal-timedata,almostforeverything,throughtheuseofradiofrequencyidentification(RFID)tools[20].ToremainprofitableinanRFID-drivenworld,industrymust “adaptordie” [21].Real-timeconsumer-driven supplychain [22]dynamicsalsodeterminethespeedatwhichtheindustrymustchangetoremaincompetitive[23].Hence,businessmodelsarecontinuouslyfocusedondata,newtechnologies[24],costreduction,andthequestfornewgrowthwithoutsacrific-ing quality of product and/or service. Systems that help to maintain “everydaylowcost” atWal-Mart andefficiencies atDell that still allow“makingboxes” aprofitable pursuit deserve strategic exploration and integration of germane ideastoimprovesustainablehealthcaredelivery.Healthcaresystemswerenotdesignedwithscientificprinciplesinmind[25],andtheethosof“innovateordie”[26]isnotsalienttohealthcareproviders.

Whilesoftwaresystems,likeenterpriseresourceplanning(ERP),generateben-efits for business and industry, through some degree of integration of data andautomationofplanning,therearefewhealthcareelectronicmedicalrecords(EMR)systems (EMRS) that have autonomous decision-making capability. Healthcare,evenpreventativeorwellness,ifavailable,isstilldependentonexpensivehumanresources for data acquisition, monitoring, analysis, reporting, and follow-up.Reengineeringthiscoststructureandhospital-centricservicemodelisnecessary.Configuringa reliablebusiness servicemodel forhealth servicesmaymeetwithentrenchedresistancetochangebutmaycatalyzeextendedhealthcareforbillionsandmayevenimprovehealthcarequalityofservice.

K10373_C008.indd 249 10/8/2010 6:14:41 PM


8.2 Problem Space8.2.1 BackgroundThemultitudeofproblemsandissuesinmanagementofhealthcarearebeyondthescopeofanypaperorbook.Here,Ibrieflytouchupononlyoneissue:acquisitionofmedicaldatatoimprovehealthcare.However,thefundamentalnatureofthisissueforcesustoaddressaseriesofintegratedproblems,someofwhichmaybetextbookcasesinbasicphysiology[27]andbiochemistry[28].Therefore,itiswellnighimpos-sibletodosufficientandequaljusticetoalltheinterdependentprocessesandareas.

Thenaivethrustofthisissueistoreducehealthcarecostbutwithconcomitantexpansionofimprovedhealthcareservicesforbillions.Healthcarecostreductionisahackneyedtopicofdiscussion,butthethrustofthischapteristosuggestsolutionswhere emerging ideas and innovation may help expand and improve healthcareserviceatacostthatmaybesoonsustainableevenbythedevelopingcountriesoftheworld.Hence,thesolutionspaceshalldealwiththeissueinfocus:innovationinacquisitionandanalysisofmedicaldata.Itisobviousthatonecanremainobliviousofthefactthatinnovationindatacollectioncallsforinnovationintoolsfordatacollectionaswellasanalysisofdata,toextractinformationandknowledgethatcanaddvaluetohealthcareservices.Duetothelatter,theproblem(andsolution)spaceofthischapterisboundtoevolveinmultipledirections,eachindicatingafurtherlineofinnovation.

OnegeneralprobleminglobalhealthcareisduetothemimicryoftheWesternmodel focused on acute-care hospital-centric view of what health “care” is sup-posedtodeliver[29].Theaphorismthat“betterhealthisinherentlylessexpensivethanworsehealth”isequallyapplicableintheWestandtheEast[30]yetseldompracticed.Infact,theprofitabilityoftheacute-carehospital-centricWesternmodelappearstobethepreferredlineofhealthservicedelivery(Figure8.1).Itreasonsthattheword“care”mustbeomittedfromhealthcare[31]becausethehospital-centricrevenuemodelisatoddswiththe“care”thathealthservicesareexpectedtodeliverforanindividualinapatient-centricviewofpersonalizedhealthcare.

Ihastentoaddthatingeneral,theacute-carehospital-centricmodelstilloffersappreciableserviceswhenitconcernsaccidentsandemergencies(A&E).Itisvitaltorespondtothechallengesofuncertainty inhealthcarestemmingfromA&E.But,thecriticismsurfaceswhentheA&Emodusoperandiisextendedtootherareasofnon-A&Ehealthcare.ThesystemicefficienciesnecessarytorespondtoA&Esitua-tionsmustcontinuetobesupported.Suggestionsinthischapterorelsewhereaboutpatient-centricpersonalizedhealthcaremustnotbeviewedasareplacementbutasarealignmentoftheexistingsystemthatisnecessaryforeconomictransformationtokeephealthcaresustainable.A&Eandnon-A&Eapproachesarenotmutuallyexclu-sive,andthereisaneedforelementsofbothsystemstocoexistformutualenrichment.

Anotherproblemthathasevolvedoverthepastfewdecadesconcernsaninabil-ity or incongruent response of the healthcare community to adopt advances in

K10373_C008.indd 250 10/8/2010 6:14:41 PM


informationtechnologyinordertodevelopanintegratedsystemsapproachtoser-vices.Thehuman-drivendecisionmodelpracticedbythemedicalcommunityisadeptinmaintainingand/orsequesteringdata,information,andknowledge,aboutcasesandpatientsinpaper-drivensilos.Itpreventstheimprovementanddevelop-mentofa learninghealthcaresystembasedoninsightandexperience.Criticsofinformationtechnologywithinthemedicalcommunitywilldefendthisstatusquobypointingto issuesofconfidentialityofpatient–doctorrelationship,privacyofpatientdata,lackofstandardsformedicalsystemsinteroperability,andquagmireofethicalguidelinesformedicalknowledgediffusion.Thecriticsare justifiedintheirclaims.ButtheyalsoremainrefractivetoextractandusetheprinciplesthathaveincreasedprofitabilityintheservicesindustryandcatapultedbusinessessuchasGE,P&G,Nokia,andWal-Marttoluminousheightsofprofitability.Inaddi-tion, thechasmbetweenmedical educationandengineeringeducationcreates alackofawarenessoftheadvancedtoolsandinformationtechnologiesthatexistandthepotentialforinnovationinintelligentdecisionsciences,inordertoprovidedataprotectionthecriticsdemandandthepatientanddoctordeserve.

Thebusinessservicesapproachtohealthcare,admittedly,raisesalarmsifindi-vidualsorpatientsaretobeviewedascost-centerswiththeadministrationtryingto functionas aprofit-center.Thisviewassumes a literal translationofbusinessservicestohealthservice,whichisnotwhattheproposalcallsfor.Asanexampleofsuccessfultransformationofbusinessserviceefficiencies,onemayciteeducation,adomainwithdistinctlydifferentdynamicsfrombusiness.TheOpenUniversityin

100%

80%

60%

40%

20%

0%

Am

bula

tory

surg

ery

Radi

olog

y

Out

patie

ntth

erap

y

Labo

rato

ry

Reha

bilit

atio

n

Pedi

atric

s

Wel

lnes

s

Obs

tetr

ics

Emer

genc

yro

om

Pro�tability of service lines

Figure 8.1 Reporting service lines as profitable (above 80% reporting) or unprof-itable (below 80% reporting). (Adapted from Deloitte Consulting, The Future of Healthcare: An Outlook from the Perspective of Hospital CEOs, New York, 2005. With permission.)

K10373_C008.indd 251 10/8/2010 6:14:42 PM


UnitedKingdom[33]andUniversityofPhoenixintheUnitedStates[34]havepio-neeredprofitablebusinessservicetypeapproaches,integratedwithextensiveuseofinformationcommunicationtechnologies(ICT),todelivereducationofareason-ablestandardforvastnumberofindividuals,whowereunabletoaccesstraditionalhighereducation,foravarietyofreasons.

Remoteaccess toeducationhadhumbleorigins in“correspondence”coursesbutwastransformedbythegrowthoftheICTsectorandacceleratedbythediffu-sionoftheInternet.Thisisaformof“personalized”educationthatenablesindivid-ualstomoveaheadincareersoftheirchoiceandcontributetoeconomicgrowth.Thiseducational–economictransformationmayofferaparadigmforthehealthcareindustry.Ofcourse,theacademiclevelofOUandothersimilaroutfitsmaynottrainanindividualtobethenext“bigbang”theoristorleadresearchertogarneraNobelPrize.Butforthoseelitepurposes,theacademicsystemiswellpreparedwithitsselectinstitutions.Theeliteacademicinstitutionsmaybeanalogoustothe“elite”equivalentinhealthcare,whichmaybetheacutecareandA&E.However,theelitemodelinhealthcaresystemmayexcludeorrestrict,ongroundsofprofitabilityorcost,theaccessibilityofnon-A&Etypepreventativeorearlydiagnosticmodesofhealthcare(analogoustoOUandsimilaroutfits)forthemasses.

8.2.2 FocusToreiterate,thefocusofthischapterrelatestoinnovationinacquisitionandanaly-sisofmedicaldatatoimprovehealthcarebyexpandingcoverageforthemassesyetdelivergreatervalueofhealthcare services.Serviceorientation, systemsarchitec-ture,andtheuseofsoftwareasinfrastructuredependondatasourcesandanalyticsneeded from healthcare monitoring, sensing, and responding to situations. Thebusiness services conceptofdata, information transparencybetween systems, aswellasdataexchangeand/orinteroperabilityissuesmaybemorecomplexinthiscontextdue to regulatory and security constraints. Itmaybequiteuseful if thebusinessservicetypeapproachcanalsointroducesomedegreeofautomateddeci-sionmaking, evenbasedon rules andworkflow, fornon-exceptional healthcarecasemanagement.

The next level of decision making based on acquired data with reference tostandards(e.g.,pulserate,bloodpressure,andnormalrangeofbloodglucose)mayrequiresomebasicalgorithmsbasedonsimpleartificialintelligence(AI)principlesthatcaninducealearninghealthcareapproachwhenevaluatingdataaboutaspe-cific individual or patient. For example, if the individual is otherwise “normal”evenunderahighersystolicordiastolicbloodpressure(BP)reading,thentheana-lyticalalgorithmcanlearnthatthedeviationfromthemedicalstandardreferencemodel(120/80mmHg)isnotreadilyacauseforalarminthisspecificcasesincethe individual isphysiologically “normal” evenunder an elevatedor lower thanstandardBPreferencedata.Hence,billionsoflearninginstancesarenecessaryforaglobalmodel.

K10373_C008.indd 252 10/8/2010 6:14:42 PM


Thisvery importantdecision, toconcludean individualorpatient isnormaldespiteaslightlyaberrantstandarddata,mustbemade,undermostcurrentcir-cumstances,byatrainedmedicalprofessional.Thattranslatestocost.Aggregatedovernumbersofinpatient-andoutpatient-relateddata,thesecumulativecostssoonbegintodestabilizethefinancialinfrastructure.Equallyandperhapsmoreimpor-tantisthetimespentbythetrainedmedicalprofessionaltoreviewthedataandarrive at the decision. Time spent for non-exception management siphons awayvaluabletimefromexceptionmanagementandpatientswhoindeedneedattention.Therefore, it isnotdifficult tocomprehendthat smallchanges in thehealthcareprocessposeminimalrisk,quantifiableusingbusinesstools(seeRef.[201]),yetmayimprovequalityofserviceandreducecost.

Documentingtheacquireddata,forexample,thebloodpressurereadingmen-tionedabove,isthenextlevelthatdeservesexplorationduetothecascadeofeventsthatthisdatamaytrigger.Withafewexceptions,eveninthemostadvancedindus-trializednations,paper-baseddocumentationisthenorm[35].Unlesspaper-baseddocumentationistheexception,ratherthanthenorm,healthcaresystemsmaycon-tinuetobecrippledfromdisplayingtheirtruefunctionalpotential.Givenhumanerrorsofdata input, thetransformationofpreexistingandaccumulatingdataaswellasnotesanddecisionsposesamajorchallenge.Withouttheavailableinfor-mationonexistingpatientsandcases,theabilityofdecisionsystemstoarriveatnon-exception-management-relateddecisions,ontheseexistingpatientsandcases,maybeseriouslyflawedandhencemayberenderedunacceptable,tohelpdealwithexistingpatientmanagement.

Thusfar,wehavereferredtopatients,butwhataboutindividualswhoarenotpatients yet? Wellness or preventative medicine and early risk identification arecriticaltoreducetheprobabilitythatanindividualshallbecomeapatientorneedacutecareoremergencyattention.Theacute-carehospital-centricmodelislargelyviewedasafailuretoaddressthisbroaderspectrumofpersonalizedhealthcareeventhoughitmaybewellequippedtodealwithA&Einnationsbig(e.g.,theUnitedStates)andsmall(e.g.,Ireland).

8.3 Solution Space8.3.1 Existing Electronic Medical Records SystemsBeforeembarkingonthediscussionofemergingtrendsandpotentialforinnova-tiontoaddresstheproblemfocusoutlinedabove(Section8.2.2),itmaybepointedoutthatmedicaldatacapturedinelectronicformat(EMRS)exist inpractice insome formor theother [36]. Itmayoffer architectural clues or serve as abasictemplate(startingpoint)fornationsbeginningtograpplewiththeproblemofcre-atingEMRS.However,expectingthecurrentEMRStoserveasa“bestpractice”orbenchmarkmaynotbeprudent.ThereisampleroomforimprovementofEMR,

K10373_C008.indd 253 10/8/2010 6:14:42 PM


which is essentiallyagenericdataaggregationplatform.Theevolutionof futureversionsorvariationsofgenericEMRplatformmaynotbearanyresemblancetocurrentsystemsthataregenerallycommand-driven,archivaldatastores,withlittle,ifany,analyticalcapabilities,suchasclinicaldecisionsupport(seeRef.[63]).

Since1907,theMayoClinic(theUnitedStates)claimstohavekeptunifiedmed-icalrecordsthatexistinelectronicformatsince1993[37]inasingledatabasewith5 millionrecordsincludingpatientfiles,x-rays,laboratoryresults,andelectrocardio-gram(ECG)records.Since2004,itincludesdataminingandpattern-recognitiontoolstodiscoverrelationshipsamongspecificproteins,geneticmakeup,andtreat-mentresponses(seeSections8.3.2and8.5.1).InformationcanbesharedbetweenthedifferentgeographiclocationsoftheMayoClinic,andphysicianscanconductavirtualconsultationonanypatientbecausetheEMRsareaccessiblefromallthreesites.Thismayrepresenttheprimordialroleofdatainpersonalizedhealthcare.

TheVistA system inuseby theVeteransAdministration (theUnitedStates)hospitalscovers150medicalcentersand1400sitesacrossthecountry[38].Eighty-fivepercentof57millionoutpatientvisitsandalmostallinpatientnotesareavail-ableonlinethroughVistA.Inaddition,94%ofoutpatientprescriptions(equivalentto200million30dayprescriptions)andalmostallofinpatientprescriptionsareenteredinVistaA(EMR)directlybytheprescribingclinician.Astudyin2004com-paredVAversusnon-VApatientsin12communitiesandfoundthatVApatientsscoredhigheronqualityofcare,chronicdiseasecare,andpreventativehealthcare.

ThescopeofbenefitsthatcanbederivedfromEMRS,asonecomponentinthe solution space, is only limited by our imagination, but, at present, severalthorny challenges remain. Unlike the Mayo Clinic records and their visibilityacrossthethreedifferentgeographiclocations,theinfrastructureofVAorPartnersHealthCare[39]ismoreextensive.Patientscanmovebetweenlocationsandtheirtreatmentcanchangeovertime.Thisintroducesmajorsystemicissues,asfollows.Partialsolutionsaresuggested,ifapplicable.

Ofimmediateconcernisuniqueidentificationofdata(seeRef.[39]).Howdoweuniquelyidentifythepatient,andpatientrecords,thatmaybeassigneddifferentIDnumbersindifferentlocationsaswellasdifferentrecordsofthesamepatientthatmaybenumberedaccordingtothesysteminoperationatagivenlocation?Thecriticalvalueofuniqueidentificationthatunambiguouslylinkstheindividualtohis/herrecords,irrespectiveofthephysicallocationofthehospital,doesnotneedemphasis.Thesheernumberofindividualrecords(laboratorytests,x-ray,CTscan,ultrasound,physician’snotes,medication,response)multipliesovertimeandmaypresentanumberingschemedilemmathatrequiresasolutionbutwithoutreinvent-ingthewheelorintroducingyetanother“new”system.

Creatinguniqueidentificationisnotacompetencyofthehealthcareindustry.Hence,thehealthservicesmayoptforanexistingidentificationschemethat(1)canoffervastnumberofuniqueaddresses[40]thatcanbeorganizedinrelationshipsorsubclasses,(2)istrulyportable,(3)Internetready,(4)alreadyinoperation,and(5)globallypre-agreedinamannerthatcanaidadoptionbythehealthcareindustry.

K10373_C008.indd 254 10/8/2010 6:14:42 PM


AproposalthatcanaddressthisissueofuniqueidentificationofoctillionsofitemsusingthegloballyagreedIPv6formatispresentedinaseparatepaper(seeRef.[40]).SupportforthepotentialuseofIPv6-basedidentificationmaysoongainmomentum[41],andtheneedforthisapproachinhealthcarewashighlightedinproblemsdiscussedelsewhere(seetestimonythatfollowsfromRef.[38]).

Personalizeduniqueidentificationinhealthcaremayofferarobustidsolution,butitsvalueforthepatient,whomaymovebetweenvarioushealthcareunits(e.g.,physiotherapy,strokecare,outpatientclinic)orhospitals,localand/orglobal,shallremainconstrainedwithoutinteroperabilitybetweensystemsofdifferenthealth-careproviders,publicandprivate,engagedwiththepatient.Thereislittlevalueindeployinguniqueidentificationifanindividualpermanentlyresidesinonelocationand receiveshealthcare fromonemedicalprofessional, inoneclinicorhospital,whichisentirelyself-containedinallitsservicesandwithoutanyexternalinterac-tion,guaranteeingtotalqualityhealthcarefortheentirelifecycleoftheindividual.

Interoperabilityseguestotheissueofstandards.TheIPv6standardgoverningtheunique identificationschemementionedabovewillmake itpossibleto iden-tify theuniquenumber(“address”)inanyhealthcaresystemanywhereintheworld,becausethestandardhasmadeprovisionsforassigningthatnumberoraddresstothatrecord,orpatient, inamannerthatshall remainuniqueoverthe lifeofanindividual.Itislogicaltoanticipatethata“numberre-claimingscheme”thatmaybedeployedtoclaimbackandreusedormantnumbers(e.g.,apatientnumbermaybeclaimedbackandreusedevery150yearsifitisassumedthatahumanbeingisunlikelytoliveformorethan150years).

StandardsshallprominentlyfeatureinEMRSsolutionsspacewhenandifthehealthcaresystembeginstoexperimentwithrule-basedorintelligentdecisionsci-encestohelpevolveclinicaldecisionsupport(seeRef.[63]).Standardsor“rulesofoperation”areimmediatelynecessaryforthegranularanddiversequalityofrecord-keepingthatcommenceswithpatienthistoryandphysicalexamination.Itismostoftencarriedoutbythelocalprimarycarephysicianorgeneralpractitioner(GP),incommunitieswhereprimarycaremaybeinoperationandwhereEMRSmaybeavailableforrecordkeeping.Ofcourse,thesamecouldcommenceforindividualswhoreporttotheA&E.

Standard“history”datainanEMRSthatisalsointeroperableareacomplexproblemthatmustdigdeepandwideformultidisciplinaryconvergencetogenerateaworkingsolution.It iscomplicatedbythemultitudeofdescriptivesyntaxthatmaybeusedbythepatientandthewrittenforminwhichthemedicalprofessionalscriptstheinformation.Thenaturallanguage(mothertongue),cognitiveabilities,culture,education,values,andexperienceofboththepatientandthephysicianormedicalprofessionalshallcolorthecontentandcontextofthishistorydocument.ItisapparentthatexistingICTandcomputersystemsmayfinditdifficult,ifnotimpossible,toextractwithreasonableprecisionthe“meaning”ofthehistorywrit-teninwordsandtransformthemtoastandardformatthatismedicallyrelevantforEMRSinteroperability.Evenmorecrucialisto“understand”theimportance

K10373_C008.indd 255 10/8/2010 6:14:43 PM


ofthephenotypicinformationrelativetothecontextofthepatientandhisorhergenotype,which,tobeofvalue,mustrefertoareferencemodelofthecommunity,geography,andenvironmentwherethepatientlivesorlived.

Totheserendipitousreader,itmaybecomeobviousthatthetransformationfromapaper-basedlocalizedhealthcaresystem,withalimitednumberofpro-fessionals from a homogeneous background, to EMRS, may help to deliverglobalhealthcareservices,byanyone,themyriadofdiverseprofessionals.Butit requires substantial researchand innovation todevelop standards [42].Thedevelopmentofthesestandardsrequiresintegrationwithcognitionandseman-tictheory.Iwillreturntothis issuewhenIdiscusstheproposalofmolecularsemantics(Section8.4).

Before“boilingthebiomedicalocean”incognitionandsemantics, itmaybeuseful if the healthcare system may globally agree to partially address the gulfbetweensyntaxandsemanticsina“quickanddirty”approach,asadoptedbybusi-nessprocessesinglobalorganizations[43]ortheeffortsspawnedbythesemanticwebmovement [44] to create biomedical ontologies [45].These solutions, albeittemporary,ifdeployedandimplementedinEMRtypesystems,arelikelytooffersome benefits, if the questions and content are less descriptive and are alreadyexpressedinamannerthatismedicallyrelevant,irrespectiveofthe“backgroundandculture”oftheinvolvedmedicalprofessional.

Thedevelopmentof“quickanddirty”rulesandpartialstandardsmayhelpwith(1)exchangeofclinicaldata,(2)definingcategoriesorcircumstanceswhenaphy-sicianinonehealthcareorganizationcanchangeoramendtheproblemlistentryof aphysician in anotherorganization, (3) conditionsunderwhichclinical stafffromoneorganizationmaydiscontinuemedicationprescribedbyanotherclinicianinthesameorganizationorfromanotherorganization,and(4)typesofdataandinformationthatmustbesecuredbyprivacypoliciesandtheirenforcement,sothatconfidentialdataandpatient informationcannotbe sharedwith,or released to,anyexternalnonmedicalorganizationwithoutdueauthorizationfromthepatient.

Finally,EMRS,evenwithitscurrenthealthcareservicehandicapsandrestric-tions,canstillprovidebusinessvaluethatmaytranslatetocostsavings.Becauseorganization-specificEMRSlikeVistAoperatedbytheVAcoveramajorityofitsoperations,thesupply-demandprofileoftheoperationcanbededucedwithafairdegreeofprecision.Usingeconomiesofscale,products,andservicesmaybeboughtatabargain.VistAoffersanindicationofvolumeofpatientsthatarelikelytobeservedandthatvolumeinformationmaybeanalyzedtoforecast[46]inventoryofsuppliesnecessarytomeettheprojecteddemand.Forperishableproducts(drugs,IVfluids,food),thedesignandmanagementofsupplychain[47]ofvendorsandpartnerscanpartnerwiththebusinessoperationsunittoensureadequateinventoryofperishables tomeet“peace” timeand“war” time[48] typevolatilities,analo-gousinthehealthcaresupplychainifonecomparesnormalcourseofeventsversusepidemics,pandemics,naturaldisasters,oractsof terrorism.However, it iswellnigh impossible to stress that thekeyperformance indicators (KPI) forbusiness

K10373_C008.indd 256 10/8/2010 6:14:43 PM


operationsandinventorymanagementorpurchasingdecisionsmustbebasedondifferentoperatingprinciplesandaresignificantlydistinctbetweenbusiness[49]and healthcare (see Section 8.5.1). Nevertheless, gaining business efficiencies inhealthcareisnotanautomaticprocesssimplybecauseEMRSoffersanaggregatedviewofpotentialconsumptionofproductsandservices.Itisforthisreasonthatsomecountrieswithnationalhealthservice,whetherservicingalarge[50]orsmall[51] population, may still suffer from an inability to take full advantage of theeconomiesofscaletodrivebusinessefficiencies.

8.3.2 Changing the Dynamics of Medical Data and Information Flow

ThethesisofthischapteroutlinedinSection8.2.2selectsmedicaldataasonecon-duitthatmayofferthepotentialtocatalyzelow-cost,high-qualityhealthcareser-vices.Ananalysisofthenodesoforiginofdatainthecontextoftheirrelationshiptocost(toacquiredata)andqualityofservice(decisionbasedondata)mayformthebasisforsuggestinghowthecurrentdynamicsmaybenefitfromaparadigmshift(Figure8.2).

GPhospital

GPhospital

Data Paper Doctor Decisiondata Paper Treatment

response

Paradigm shift

ServiceHR

ServiceHR

ServiceHR

ServiceHR

ServiceHR

$

$ $

$ $

$$

$

$

$ $ $ $ $ $HR HR

Globalreference

model

Intelligentanalysis Exception

Doctor Decisiontransmit EMR

Treatmentresponse

ServiceHR

ServiceHR

ServiceHR

HR

Monitortransmit EMR

Figure 8.2 A generic model of data flow in healthcare and potential benefits of a paradigm shift.

K10373_C008.indd 257 10/8/2010 6:14:43 PM


Insomeindustrializednationsandinthedevelopingworld,thecurrenthealthcarepracticemaybelooselyrepresentedbytheschematicoutlineshowninthetoppor-tionofFigure8.2.Inthecurrentscenario,almostallpointsofinteraction(denotedbyrectangles)incurcost(denotedbylarge$signs),andeachactionconsumestimefromprofessionals(denotedby“HR”and“ServiceHR”).Thelatter,asaresult,isadrainontheavailableworkinghoursofmedicalpersonnelandisanimproperuseofvaluabletime,whichcouldhavebeen,otherwise,puttobetterusebyofferinghealth“care”servicefocusedontheelementsconnectedtothetreatmentofthepatient.

TheparadigmshiftoutlinedinthelowerportionofFigure8.2isneithernewnoraninnovativebreakthrough.Itisastrategicprocessengineeringthatisdrivenbycommonsense.Itsuggestshowbasicintegrationofsometoolsandtechnologiesmayleadtosavingsincostandimprovementinqualityofcarebyfocusingthetimeofmedicalprofessionalsonthepatient,ratherthanonstandardtasksandchores.In thisscheme,thelackofthelarge$signsunderthepointsofinteraction(rectangles)indicatespotentialforcostsavings.Abovethepointsofinteraction (rectangles),thesmall$signsimplythatimplementingandmaintainingthesechangesarenotfree,but thecost is lower(thanthe large$signs)andreturnon investment (ROI) ishigherifthecostofinstallingthesystemsisamortizedovertheirproductivelifecycle.Lesstimespentbymedicalpersonnelatvariousstages(lackofHRandser-viceHR)leavesmoretimetofocusonpatientcare.

Threeelementsthatdrivetheparadigmshift(Figure8.2)are(1)datamonitor-ing,addressed inSections8.3.3and8.3.4, (2)electronicdatacaptureorEMR,discussedinSection8.3.1,and(3)diversionindataflow.Intheremainderofthiscurrentsection,Iwilldiscussthelattersinceitisperhapsthekeyintheproposedparadigmshift.

Diversionofdataflowleadstoaninformationloopthatchannelsacquireddatatoflowthroughaglobalreferencemodelanddirectstheoutcomeofanalyticstomedicalpersonnelifitappearstobeanexception.Byconcentratingonthepatientsthatneedattention(exceptions),thesystemisabletoimproveitsqualityofservice.

Criticsareeagertoposethethornyquestion:howreliableisthemachine-basedanalyticalprocess?Duetotheexperimentalnatureofthescientificprocess,Ifinditreasonabletoconcludethattheremaynotbe,ever,aunanimouslyacceptable,complete, and absolute scientific certainty, with which anyone can predict thata machine-learning process can be guaranteed to be foolproof. Errors in medi-caldiagnosisbymedicalexpertsarerare.Thus,neitherhumansnormachinesareentirely infallible. It follows, therefore, that the “thorny” question is the wrongquestion.Attemptstofindthepreciseanswertothewrongquestionhave,thusfar,fracturedthedeterminationof,andseriouslydistracted,thehealthcareanddeci-sionsciencesexperts’ (seeRef. [63])effort, to focusonfindinganapproximatelyreasonableanswertotherightquestion:howmuchandforwhattypeofcasescanwegenerallydependontheanalyticaltoolsinclinicaldecisionsupport?

IntheUnitedStates,healthcarespendingwasinexcessof$2.1trillionin2007andprojected todouble to$4.2 trillionby2016.Thus, reducing10%ofhealth

K10373_C008.indd 258 10/8/2010 6:14:43 PM


servicesworkload throughmonitoringandanalyticsmay save theUnitedStatesnearlyhalfatrilliondollars[52]inafewyears.AsmallcountrylikeIrelandmaysaveacouplebillioneuro[53]peryear.Itmayalsotranslatetoa10%improvementinthequalityofservice(QoS).Financialsavingsmaybereinvestedincommunityprimarycarecenters,homehelpforindependentliving,andearlyriskidentifica-tion,forexample,fordiabetes.

Therefore,theinformationloopproposedintheparadigmshiftmustbecre-atedandimplemented,soon,eveniftheconstructionoccursinstepsormodules.Inevitable“growingpains”areexpectedtoaccompanyanychangeofdirection.The“loop”isagenericexpressionthatinvolvescriticalsub-elementsofgreatdepththatrequireprecisionknowledge.Thischapterdoesnotprovideaprescriptionforbuild-ingthesub-elementsbutstrivestopresentexamplesthatmayconveythenatureofthesecomponents.Itgoeswithoutsayingthattheloopisnotan“IT”jobbutdemandscollaborationbetweenITandmedicalexperts.

Figure8.2alludestoaGlobalReferenceModel(GRM)wheretheacquireddatais fedviaadatabase, suchasanEMRS.Theschema indicates that thedataflowthroughtheGRMintotheintelligentanalyticsdomain.Whatisunclearfromtheillustrationandrequiresexplanation is that thesuggestionpositions theGRMtorepresentanumbrella,orcollection,ofmultiplemodulardatabases,eachembeddedwithsomeformofrule-basedanalyticalenginethatcanevaluatetheincomingdata(streamingdata)anduseconventionalworkflowtoqueryitsdatabase(specifictothatmodule)tofindrelationships,homologies,ordiscrepanciesbasedonitsownstoreddata,specificallywithreferencetoandinthecontextofthatmoduleoftheGRM.

Forexample,apatientwithelevatedtemperaturesufferingfromsevereboutsof coughing undergoes exploration in an outpatient clinic. A nurse records thebloodpressureandtemperature,drawsbloodfortotalbloodcount,and,basedontheethnicityofthepatient,decidestotakeasputum(saliva)sampleinadditiontoadministeringaManteauxtest.Itisanimmunologicaltestdesignedtodetecttuberculosis(TB),aninfectiousdisease[54],causedmainlybythemicroorganismMycobacterium tuberculosis.

WhattypesofmodulesintheGRMcanprocessandanalyzethedatafromthispatient?MedicalexpertscandefinethenatureoftheGRMsub-modules,but,forthenonmedicalreader, itmaybeof interesttonotethefollowing.Referencefornormalbloodpressure,correlatedtoage,iscommon,asistemperature.TotalbloodcountisacommonstandardandiseasilyincludedinaGRMmodule.However,the GRM module that can deal with the results of the Manteaux test requiresmedicaldetailsaswellasenvironmentaldetailsthatrelatetotheepidemiologyofTB,lengthofhabitatofthepatientingeographieswhereTBisprevalent,immu-nologicalprofileofindividualswithconfirmedinfectionbyMycobacterium tuber-culosis.TheManteauxtestisanintracutaneoustuberculinskintestusuallyappliedontheforearmandcontainstuberculinpurifiedproteinderivative(PPD)toelicittheimmuneresponsethatisvisibleontheepidermiswithin2–3days.ForapatientwhowasbornorlivedinSoutheastAsiaoraresidentofwarmhumidcoastalareas

K10373_C008.indd 259 10/8/2010 6:14:44 PM


in the United States, a positive Manteaux test, but measuring less than 10mmintransversediameterof induration,asdetectedbygentlepalpationat48–72h,isnot indicativeofTBbutrather indicates tuberculinhypersensitivity, resultingfromcontactwithnonpathogenicenvironmentalmycobacteriaorchildhoodvac-cination by Bacillus Calmette-Guerin (BCG), which is an attenuated strain ofMycobacterium bovis[55].

TheaboveexampleillustratesthewebofrelationshipsthattheGRMandana-lyticsmustbe able to extract and transmit, to thepoint-of-caremedicalprofes-sional,usingavisualizationinterfacesuchaspersonaldigitalassistantorBlackberrytypemobilephone.However,irrespectiveoftheapparentcomplexityoftheaboveexampletothenonmedicalreader,mostofthedataandinformationmentionedabovearealreadyavailableinseveraldatabasesandareclassifiedundervarietyoftopics, includingtheobviousheadingof infectiousdiseases.There isnoneedtocreate,denovo,anybasicmedicaldataorinformationdatabase.Forexample,fromthescenarioabove,thesputumsamplefromthepatientmayholdseveralcluesforearlydetectionanddiagnosis,basedonadvancesinsaliva-basedbiomarkers[56].ThedatafromsputumanalysismaybetransmittedtotheGRM,anditcanquerytheSKBorSalivaomicsKnowledgeBase[57]toextracttheinformationforfurtheranalysis.Severalsuchdatabasesexistwithspecializedknowledgeandinformation,whichareaccessibleviatheWorldWideWeb.Unfortunately,thetraditionalwebworksasadirectedgraphofpageswithundifferentiatedlinksbetweenpages.Thisis not conducive to the type of relationships necessary for healthcare analytics.Emergingprinciples fromsocialnetworkingmaybequitehelpful forhealthcareserviceanalytics(seeSection8.5.1).Blogosphere(seeRef.[10])hasamuchrichernetworkstructureinthattherearemoretypesofnodesthathavemorekindsofrelationsbetweenthenodes.Deployingtheprinciplesofblogosphereinhealthcareanalysismaybequitepromising.

Thus,thechallengeistofindnewwaysortoolstoidentifyandrelatetheselectedsourcesthatmayserveascomponentsofGRMthroughavirtualamalgam.GRMrequiresamechanismtosearchanddetecttheinformationdatabaseandthenquerythedatabasedependingonthecaseorpatientunderinvestigation.Foreverypatient,thesestrandsofdata-dependent,symptom-dependent,ortest-dependenttasksmustbecreated,inrealtime,ondemand,perhaps,asahigherlayerintegratedabstractionintheformofanapplicationmodule(poorchoiceofwordbuthopefully,itconveystheconcept).Forexample,continuingtheTBscenario,thedatafromtheManteauxtestplusthesymptomofcoughandtheeosinophilcountfrombloodtestmayserveasthreevariablesthatmaytriggeranadhocapplicationthatasks,eitherassinglequeriesorcollectively,“WhatisthepotentialdiagnosisiftheManteauxtestrevealsa10mmtransversediameter,chestycoughsarepersistent,andeosinophilcountis8%?”Toexecutethisprocess,thesystemmaycreateanadhocapplication-specificinterface(ASI),application-specificquery(ASQ),andanapplication-specificrela-tionship(ASR)thatcanact,eitheraloneorasabundledapplication,toproberelevantknowledgerepositoriesordatabases, to extract the information.This information

K10373_C008.indd 260 10/8/2010 6:14:44 PM


maybefurtherrefinedbytheintelligentanalyticscomponentintheinformationlooportransmittedtothepoint-of-care(POC)medicalprofessional.

Thescenarioabovemaybe looselysuggestiveofamedicalexampleofmash-upsoftware[58] that isgainingpopularity inbusiness servicesutilizingSOAorservice-orientedarchitecture[59].SOAistoutedtohelpbusinessservicesremainagileandadaptabletomeetthecompetitivechallengesduetovolatilityofconsumerpreferences and uncertainty stemming from globalization of the supply chain.To capture sales,thesebusinessservices,ondemand,inrealtime,createapersonal-izedwebserviceanddisplayacollageofobjects,optimizedforconsumerchoice.The collage is culled from different domains or databases that may or may notbelongtothebusinessbutsecurecontentondemandthroughlicensing.Themash-up appears seamless through the wizardry of visualization tools. The consumerviewsitasawebpageonacomputerscreenormobilephone.Theviewmayinclude,forinstance,acompanylogo(asoftwareobject),priceofaproduct(databasetableformat, stored as an object) aimed at a market segment (extracts clients prefer-encesandmatcheswithstoredclassesofobjects,e.g.,sports),andorderinginforma-tion(anotherstandardshippingandhandlingobject)withlinkstotrackandtracedetailsprovidedbyathird-partylogisticsprovider(e.g.,linktoFedExsite).

Theanalogyintheabovetwoparagraphsmayfallshortofthespecificsorlacktheprecisionthatexpertsinrespectivedomains(medicineandinformationtech-nology)maydemand,butitmayconveytheessenceofcase-specificexploration,ondemand,integratedtoknowledgediscoveryfromdatabases,orotherdomains,nec-essaryfordeliveringhealthcareanalytics.EachcountrymaycreatetheirownGRMinfrastructuretooptimizehowtheGRMmayberelevanttothenationanddelivervalueinmedicalanalytics,atleastincasesthataresimpleenoughtobeacteduponbymachineintelligence,wherereasonableconfidencecanbeplacedinthedecision.Determiningwhatis“simple”mayvary,widely.

Triggeredbypatientdatainput,thesearchfunctionofGRMmayevokethenotionofMedicalGoogle.OnedifferenceisthatthesearchisnottheendpointofGRMbut is for theGooglesearchengine[60].Thegranularityof thesearchprocessimplicitinGRMalsodiffersfromthatofGoogleinitsquestforknowledgedatabases, followedbytheextractionofrelevantdataand/or informationand/orknowledgethatmustbefirst“discovered”andthenresynthesizedandpresentedtotheintelligentanalyticalengine.

The“intelligentanalysis”referredtoinFigure8.2isa“placeholder”formultipleanalyticaltoolsthatareavailableandmaybedevelopedinfuturewherethe“learn-ing”abilityofthetoolsmaybeakeyemphasisinadditiontorule-basedapplica-tions.Thetools,forexample,artificialneuralnetworks(ANN),originatefromthedomainofartificialintelligence(AI),andnewalgorithmsmaycontinuetoevolve.Sincemedicineisanintenselyintegratedscience,thenetworkofinterrelationshipsbetweenmedicalparametersandphysiological function iskey tounderstandinghealth.Theplethoraofreasonsthatmayoffergenericsymptoms,forexample,fever,makes it imperative that thepoint-of-care (POC)physician is sufficiently aware

K10373_C008.indd 261 10/8/2010 6:14:44 PM


ofthespectrumofreasonswhyanindividualmaypresentthesymptomsoffever.Presentationofalistofreasonsmaystemfromtheuseofrule-basedenginesthatmaysearchandcompilealist.However,thevalueofthe“list”islimitedunlessthecontextofthepatientandhistoryistakenintoconsideration.

Rules,tocombineandselectthebestpossiblematchbetweenthelistandcon-text,maybecreated,buttherigidityofrule-basedselectionmaymakeitlessreliableifcomparedtoasetofalgorithmsbasedonAIprinciplesthatcan“learn”andforgetthesubtle,ornotsosubtle,changesinthecontextofthepatientthatmayincludeparameters,suchasage,activity,profession,environment,habitat,andnutrition.Forthisreason,GRMandtheintelligentanalysisclusteroftheinformationloopmayusedatacubes[61]andcomponentsthatmaybecountryspecific,nationspe-cific,orcommunityspecificbutmayalsodrawonsynergiesbetweendemographi-callyrelatedco-localizednations(e.g.,ScandinaviaorEurasianSteppes).

LetusconsideranexamplewhereJane,11,contractsafeveronMondaymorn-ingafterahotweekendthatshespentonthebeachandswimmingintheocean.OnMonday,Patrick,71,complainsoffever,too.Healsohaschronicobstructivelungdisease(COLD).TheattendingphysicianmayfocusondeterminingwhetherJanemayhaveanearinfectionwhilegearingtotreatPatrickforchestinfection.Butcanweapproachthislevelofinteractionandperhapsatreatmentsuggestionwithoutincurringthecostandtimeofthephysician?Theuseofacquireddata(seeSection8.3.3)anddecisionsciencesmayofferaroutetosavings.Toexecutethisinteractionandofferareliable,low-riskdecisionortreatmentsuggestion,intelligentanalyticsviewstheGRM-evaluatedrawdatapluslistofpossibilitiesforthefeverandintegratesthecontextwithpatienthistory.ItalsocheckstheGRMpharma-cologymodule.Thesystemtransmitstheexploratoryanalyticalsequenceloganddiagnosisandrecommendsage-appropriateantibiotics,ineachcase.

Although the use of artificial neural networks (ANN) and artificialintelligence-basedalgorithms(e.g.,ant-basedalgorithms)wasmentionedonlyforbusiness services applications (e.g.,mash-up), theuse ofAI-based agent systems[62]canprovidearobustandgranularsystem.TheexperimentaluseofAIinclini-caldecisionmaking[63]andproductive implementationofAI in industry[64],business[65],business-to-businessexchanges[66],andotherapplications[67]pro-vides confidence that agent-based systems may become the norm in healthcare,too.Duetotheinterrelatednatureofmedicine,numerousparametersmustbecon-comitantlyevaluatedforanydecision,andeachpatientmustbetreatedasaninde-pendentinstancethatisspecificforthatpatientonly.Datarelatedtothepatientalwaysremainpatientspecificwithoutsharing,aggregating,orclusteringdata,inany form, whatsoever. For each patient, the classes of data and volume of datapointsarelikelytobequitehigh.Alldatapointsmustbestoredandrelationshipsevaluatedfordiagnosisandprognosis.Therefore,theuseofdatacubesandtheabil-ityofagentsystemstoconnectbetweenalldatapointsthroughthecube-on-cubeorganizationofdatacubesmaymakethisapproachparticularlyessentialandben-eficialforthebillionsofinstancesnecessaryforhealthcareservices.

K10373_C008.indd 262 10/8/2010 6:14:44 PM


RevisitingtheearlierexamplewhereaManteauxtestwasadministered,fromanagentperspective,theapplications,inthatdiscussion,maybedissociatedintosingleagents,eachwithitsspecifictaskinrelationdataandexplorationofthewebof relationships with respect to the data or task assigned to the agent. In otherwords, theagent thatholds the resultsof theManteaux test (data=10mm) fortheobservedindurationischargedwiththetasktofindout,throughthemediumoftheGRM,theimplicationsoftheobserveddata(10mm).Thedevelopmentofintelligentagentssystems[68]iswithinthegraspofcurrenttechnology(seeRef.[182])andcanbeimplementedinhealthcaresystems.The“bundled”higherlayerabstractionof individualapplications(ASI+ASQ+ASR)mentionedearlier inthissectionisanalogoustomulti-agentsystems[69]thatworkwithagentsinahierar-chicalfashionwithhigherlevelagentstaskedtointegratetheinformationordatafromlower-levelagents.Itis likelythatmulti-agentsystems(MAS)maybecometheworkhorseoftheAI-basedintelligentanalyticsinhealthcare.

AI-basedagentsgenerallyuseprogramming languages fromtheopensourcedomain.Hence,agentsarehighlymobileandsuitedtoqueryavarietyofdatabasesforknowledgediscoveryusingopen source tools, for example,RDFor resourcedescriptionframework[70]andOWLorontologyworkinglanguage[71].Agentsarelikelytoformanintegralpartoftheemergingsemanticweb[72].But,agentsneedspecialinterfacesif interactingwithproprietarydatabases.Proprietarysoft-warevendors[73]denyRDFfromaccessingtheirdatadictionariesthroughtheuseofproprietaryprogramminglanguage,forexample,theuseofproprietaryABAPprogrammingbythesoftwarebehemothSAPAG.Theseproblemsmayspurnovelapproaches. One data transformation tool called Morpheus [74] may facilitateextractionofdata fromvarious locationsby transforming them into a commonformat,which is thensent to“holdingtank”orarepositoryof transformations.Morpheus,usedasabrowsertool,mayhelptofindarepositorytransformationthatGRM(Figure8.2)maybeseeking.ThetransformationtoolmaydraganddropdataorinformationinaformatusedbyGRM.

Inanothervein,theopensourcemovementiscatalyzingthediffusionofsoft-waretoolsthatmaymakeiteasierforagentstoaccessproprietaryformatsthroughstandardapplicationprogramminginterfaces(API)thatstillpreservethepropri-etarynatureofthesystembutthrougha“translationalinterface”orflatfiletypeformatexchangesdataor information. If thedataor information soughtby theagentissecuredandinneedofauthorizationforrelease,agentsystemsarecapableofexchangingproofstoprovidesuchauthorizationandreleasethedata.Themobil-ity of agents raises importantquestions aboutdata security and its implicationsforhealthcare.Itiswelldocumentedthatagent-basedmodelsaremorerobusttoensuredatasecuritybyvirtueoftheAIalgorithmsusedinitsconstruction.Atpres-ent,generally,mostsoftwarearchitecture,forexample,ofthetypeusedinEMRS,dependsonequation-basedmodelsthatareinherentlyfarlesssecure.

Takentogether,agent-basedarchitecturemaysoonbecomepervasiveinemerg-inghealthcaresystems.Agent-basedsoftwaremayformpartoftheinfrastructure

K10373_C008.indd 263 10/8/2010 6:14:44 PM


ofthehealthcaresystemofsystems(HSOS).HSOSnotonlyconsistsofthecompo-nentsillustratedinFigure8.2butextendstoincludemobileagentsthat(1)monitorthefunctionalstatusofmedicaldevices,(2)schedulehumanresources,(3)aidintheplanningofmealservicestomatchnutritionalneedsorrestrictionsofpatients,(4)guidebusiness functions tobenefit fromeconomiesof scale, and (5)overseefinancialrecordstomonitorthefiscalhealthofthehealthcareservice.Otherformsofconventionalnon-AIsoftware(Morpheus,mash-up,SOA,ERP,webservices)maycoexistwithintheagent-basedsoftwareinfrastructureforroutinetransactions.

Agent-based“learning”systemsmayaugmentthedepthandprecisionofdataminingandpatternrecognition(seeSections8.3.1and8.5.1).Rule-baseddatamin-ingandpatternrecognitionmaybeout-of-datesoonafter“new”rulesareupdated.The latter may be particularly relevant to healthcare tools, data structures, andanalysisofparametersindiagnosisorearlyriskidentification.Systemsmustcon-tinuouslylearn,adapt,orimprovetoextractandusethesubtlechangesthatmaybeindicativeoffuturediseasepotentialorcandifferentiatebetweencloselyrelatedtypesofanomalies.Someoftheseparametersmaydifferorbealteredsometimes,albeit slightly, betweenpopulations [75]. In the contextof thepatient, that canimpact theoutcome,considerably, toprevent falsepositivesorwrongtreatment.Agent systems can continuously “learn” about changes and hence offer greaterconfidence in the outcome of agent-based GRM-integrated intelligent analytics(AGRMIA) compared to rule-based tools. Algorithms for relation analysis areemerging from research on social networking [76] and reality mining [77] thatmaybeadaptedforuseinAGRMIA(seeSection8.5.1).

Applicationofpatternrecognition,inonestudy,achievedperfectdiscrimina-tion(100%sensitivity,100%specificity)ofpatientswithovariancancer,includingearly-stagedisease[78].Thestudyidentifiedsubsetofproteomicbiomarkersusingmassspectroscopyofproteomicanalysisofserumfromovariancancerpatientsandcancer-freeindividuals.Statisticalalgorithmsanalyzedthemassspectraldataandselected,usingrandomfieldtheory,allbiomarkersthatweresignificantlydifferentinexpressionlevelsbetweenaffectedandunaffectedsubjects.Thebestdiscriminat-ingpatternwaschosenamongallsignificantbiomarkersbyusingthebest-subsetdiscriminantanalysismethod(LinearDiscriminantAnalysis).Anotherstudyalongthe same lines developed an algorithm employing principal component analysisfollowed by linear discriminant analysis on data from mass spectrometry andachievedsensitivity,specificity,andpositivepredictivevaluesabove97%onthreeovariancancerandoneprostatecancerdataset[79].DetectionofovariancancerusingsensitivemolecularbiomarkersisespeciallyurgentinwomenwhohaveahighriskofovariancancerduetofamilyorpersonalhistoryofcancerandforwomenwithageneticpredispositiontobreastcancerduetoabnormalitiesingenessuchasBRCA1andBRCA2[80].

Applicationof remotemonitoring (seenext section)ofbodyfluidsusingpro-teinmicroarraychips[81]thatcantransmitdata,advancedmathematicaltoolsforbiomarkerdataanalysis,AI-basedintelligentpatternrecognition,andagent-based

K10373_C008.indd 264 10/8/2010 6:14:44 PM


GRMinformationflow(AGRMIA),iftakentogether,mayholdpromiseforglobalhealthcare.Tools,suchasmassspectroscopy(MS),providecluesaboutmolecularidentitiesofdifferentiallyexpressedproteinsandpeptidesinbodyfluidsorinbreath[82]thatmaybecriticalforearlydiagnosis.Agent-basedsystems,operatingthroughtheGRM(Figure8.2),canextractthesetypesofdataaswellasinformationcata-loguesofbiomarkersfromotherfields[83]andapplytheknowledgetopatientdata,to identify risk and improvediagnosis.Hence,MSanalysisofproteinorpeptidebiomarkers[84]inbodyfluidsusingmicro-fabricatedminiaturizedMSdevice[85]operatingasalow-costwirelesssensormayofferageneralpopulation-basedassess-mentofproteomicpatterntechnology,asascreeningtoolforearlyriskidentificationfor several diseases, to complement lab-on-a-chip type sensors for earlydetectionof cardiovascular diseases (CVD) [86] and carcinomas [87]. A proposed systemsapproach,bywhichmassspectroscopicdata(proteinandpeptidebiomarkers)maybecomparedbetweensystems,willbeexploredinSection8.4onmolecularsemantics.

8.3.3 Data Acquired through Remote Monitoring and Wireless Sensor Network

TheparadigmshiftinFigure8.2illustratesthatdataacquisitionandtransmission,evenifpartiallyassistedbytheuseofmedicaldevicesforremotemonitoringtoolsandinformationcommunicationtechnologies,mayreducecostandfreeuptimeformedicalprofessionals.Inprinciple,fewcanargueaboutthevalueofthisapproach.InSection8.3.2,referencesweremadetopotentialforremotemonitoringandsen-sorstoimprovehealthcareservices.Inthissection,Ifocusononeremotemonitor-ingdevice.Thebasicstrategy,fromamedicaldeviceperspective,maybesimilarforthemajorityofvitalmeasurements(data)carriedoutbytheprimarycareGPoratthehospital.Securityoftransmitteddataandunauthorizedaccessisprevent-ableusingagents.Toguaranteeevenmorestringentdatasecurity,recentresearchonPUForPhysicalUnclonableFunctions[88]maygenerateunique“fingerprints”thatcandistinguishidenticalchipsorICfromthesamemanufacturingbatch,thatareusedinbio-sensorsandothermedicaldevices.

Remotesensingtechnologiesarewelldeveloped[89],yet theirapplicationtononinvasive,wearablebio-instrumentationcapableofwirelesstransmissionofreli-abledatahasonlyemergedinthepastfewyears.Oneinnovativedevice,theRingSensor(Figure8.3),hasemergedfromtheconvergenceofrobust self-organizingwireless radio frequency (RF) transmission and an improved photoplethymo-graphic(PPG)wearablesensortomonitorvitalsigns[90].TheRingSensormini-mizesmotionartifactswhenmeasuringarterialbloodvolumewaveformsandbloodoxygensaturation,noninvasivelyandunobtrusively,fromthewearer’sfingerbase.Figure8.4showstheresultsfromRingSensormonitoringofheartrate(datatrans-mittedthroughawirelesssensornetwork)andcomparestheresultstoconventionalelectrocardiogramandwiredfingerphotoplethymograph.Thelatterissusceptibletomotionartifacts.

K10373_C008.indd 265 10/8/2010 6:14:44 PM


Circuit boards

Batteries

Inner ring(bastic fabric)

Outer ring(aluminum)

Figure 8.3 Noninvasive ring sensor for wireless monitoring of heart rate. (From Rhee, S. and Liu, S., An ultra-low power, self-organizing wireless network and its applications to non-invasive biomedical instrumentation, in IEEE/Sarnoff Symposium on Advances in Wired and Wireless Communications, West Trenton, NJ, March 13, 2002. © 2002 IEEE.)

1101009080

1101009080

1101009080

0 10 20 30 40 50 (sec)

0 10 20 30 40 50 (sec)

0 10 20 30 40 50 (sec)

(a) Heart rate (beats/min) by electrocardiogram (ECG)

(b) Heart rate (beats/min) by fingertip photoplethymograph (PPG)

(c) Heart rate (beats/min) by wireless ring sensor shown in Figure 8.3

Figure 8.4 Heart rate monitoring by conventional, wired and wireless devices. (From Rhee, S. and Liu, S., An ultra-low power, self-organizing wireless network and its applications to non-invasive biomedical instrumentation, in IEEE/Sarnoff Symposium on Advances in Wired and Wireless Communications, West Trenton, NJ, March 13, 2002. © 2002 IEEE.)

K10373_C008.indd 266 10/8/2010 6:14:46 PM


Itdoesnotrequireanystretchof imaginationtoslipontheRingSensor(orarefinedversionofasimilardeviceshowninFigure8.3)onapatient’sfingertomonitorkeyvitalsigns,suchasheartrate,continuously.Real-timestreamingdataunderthe“watchfuleye”ofadedicatedAI-basedagent,embeddedinthemoni-toringsystemoreven intheRingSensoroperatingsystem(OS),monitorwave-formsinreal-time.ItmaybesimilarinprincipletomoteswithTinyDB(database)andTinyOS[91].IftheRingSensor“senses”reasonabledeviationofthePQRSTwaveinthecontextofthepatient(ratherthanthePQRSTstandardinGRMorglobalreferencemodel),thenitimmediately“responds”bysendinganalert(codeblue,codered)tothePDAormobilephoneofthemedicalprofessionalonduty.ReferencetocontextisimportantforpatientswithchronicCVDinordertopre-ventfalsealarms.PatientsdiagnosedwithmyocardialinfarctionoranginapectorismaydisplayaPQRSTwaveformthatmaybedifferentfromthestandardGRMversion,but thisalteredPQRSTwaveformmaybe the“patient-specificnormal”waveform.Thedatamonitoringandanalysiscomponentsof thesystemmustbeabletocontextualizethisdifference.

Thisdistinctionintheanalysisofmonitoreddatahighlightstheneedforcau-tiontotreatthissuggestionasamedicaldevicebusinessbonanza.Thechallengeistocombinethesophisticationofthewaveformmonitoringmedicaldevice(e.g.,theRingSensor)withpatient-specificcontextandhistoryunderthe“supervision”ofanagent(intelligentanalyticaltools).Agentsmayoptimizethe“sense,thenrespond”outcome, to be transmitted in a visual format comprehensible by a consultantcardiologistaswellasastudentnurse,toinitiatethemedicalprofessional-drivenresponse,decision,andtreatmentplan,ifneeded.

Inpractice,patientswithcardiovascularproblemsareoftenrequiredtowait.Intermittentmonitoringoftheirvitalsignsispossiblewhenthestudentnurseortraineegetsaroundtothepatient.Physiologicaleventsthatmayhappeninbetweenthehumanresource-dependentmonitoringmaywelldeterminethelong-termmor-bidity(strokevictims)ormortalityofthepatient.

In the at-home scenario, Ring Sensor type applications offer greater value.Continuous monitoring is the key to preventative care, in this case, for peoplewithchronicCVDorindividualsatahighriskofCVDthatmaystemfromotherconditions, for example, increasing blood cholesterol or obstructive pulmonarydiseases.Transmitteddatafromcontinuousreal-timemonitoringinthehome,ifsubjectedtoreal-timeanalysis(systemslocatedinthecommunityorprimarycarecenterorlocalhospital),arelikelyto(1)improvethequalityoflifeforthepatient;(2) reducehealth serviceexpensesbykeeping thepatientoutof thehospital forlongerperiods;(3)optimizeresourceplanningbycreatingamanagementplanandpredictingwhenthepatientmayneedtovisitGPoroutpatientclinicfornon-acutefollow-uportreatment;(4)reducehealthserviceexpensesanddemandonservicefromA&Emedicalprofessionalsbydecreasingtheprobabilityforacute-careemer-gency services thatmaybe requiredby thepatient,moreoften,without remotemonitoringprovisions;(5)improvetheoverallqualityofhealthcarebyextending

K10373_C008.indd 267 10/8/2010 6:14:46 PM


remotemonitoringnotonlyforpatientsbuttheat-riskgroup,aswell,byusingtheRingSensorwithanultralowpowerwirelessdevice,the“i-Bean,”whichisanadhoc, self-organizingnetworkprotocol that is as simple asplugging in awirelessWiFi router in home or office (Figure 8.5) or any location that can connect tothemedicalanalyticalsystemthroughtheInternet;and(6)expandthereachofhealthcareservices(withoutmortgagingthetreasury)byextendinglow-costremotemonitoringtoanotherwisehealthydemographicwhomayvolunteertokeepaneyeontheircardiovascularwellnessprofile.

The benefits from remote monitoring are undoubtedly robust, but it is alsonecessarytoremaincautiousbecause,whetherremoteoron-site,wirelessorwired,local or global, healthcare produces data that may be difficult to interpret, andlack of proper interpretation may be fatal. The role of the medical professionalandhuman-drivendecisions,evenforapparentlyroutineinstances,suchasbloodpressurecheck,may,sometimes,castadoubtifdataorsymptomsaretoogeneric.Table 8.1 (in Section 8.3.4) illustrates this point by highlighting that monitor-ingbloodpressureandrecordinganelevated, lower,ornormalreading,insome

Grid/cloud computing

Base stationi-Bean

repeater

Massachusetts General Hospital

i-Bean

INTERNET

Figure 8.5 Wireless cardiac monitor uses plug-n-play i-Bean sensor network. (From Rhee, S. and Liu, S., An ultra-low power, self-organizing wireless network and its applications to non-invasive biomedical instrumentation, in IEEE/Sarnoff Symposium on Advances in Wired and Wireless Communications, West Trenton, NJ, March 13, 2002. © 2002 IEEE.)

K10373_C008.indd 268 10/8/2010 6:14:46 PM

Future Healthcare ◾ 269Ta

ble

8.1

Sing

le G

ene

Dis

ease

s Th

at E

leva

te o

r Lo

wer

Blo

od P

ress

ure

Dis

ease

Mu

tati

on

Mo

lecu

lar M

ech

anis

mEf

fect

on

Blo

od

Pr

essu

re

Glu

coco

rtic

oid

-re

med

iab

le

ald

ost

ero

nis

m

Du

plic

atio

n o

f gen

es

enco

din

g al

do

ster

on

e sy

nth

ase

and

11

β-h

ydro

xyla

se, c

ause

d b

y an

un

equ

al c

ross

ove

r

Ecto

pic

exp

ress

ion

of a

pro

tein

wit

h

ald

ost

ero

ne

syn

thas

e ac

tivi

ty r

egu

late

d b

y co

rtic

otr

op

in; i

ncr

ease

d p

lasm

a vo

lum

e

Incr

ease

d

Ald

ost

ero

ne

syn

thas

e d

efici

ency

Mu

tati

on

s in

the

gen

e en

cod

ing

ald

ost

ero

ne

syn

thas

e

Def

ecti

ve a

ldo

ster

on

e sy

nth

ase

acti

vity

; d

ecre

ased

pla

sma

volu

me

Dec

reas

ed

21-H

ydro

xyla

se

defi

cien

cyM

uta

tio

ns

in th

e ge

ne

enco

din

g 21

-hyd

roxy

lase

Ab

sen

ce o

f cir

cula

tin

g al

do

ster

on

e; d

ecre

ased

p

lasm

a vo

lum

eD

ecre

ased

Ap

par

ent

min

eral

oco

rtic

oid

ex

cess

Mu

tati

on

in th

e ge

ne

enco

din

g 11

β-h

ydro

xyla

seA

bse

nce

of c

ircu

lati

ng

ald

ost

ero

ne;

dec

reas

ed

pla

sma

volu

me

Incr

ease

d

Hyp

erte

nsi

on

ex

acer

bat

ed b

y p

regn

ancy

Mu

tati

on

in th

e lig

and

-b

ind

ing

do

mai

n o

f th

e m

iner

alo

cort

ico

id r

ecep

tor

Act

ivat

ion

of t

he

min

eral

oco

rtic

oid

rec

epto

r b

y st

ero

ids

lack

ing

21-h

ydro

xyl g

rou

ps

(pro

bab

ly

du

e in

par

t to

the

rise

in p

roge

ster

on

e le

vels

d

uri

ng

pre

gnan

cy)

Incr

ease

d

Pseu

do

-h

ypo

ald

ost

ero

nis

m

typ

e I (

auto

som

al

do

min

ant)

Loss

-of-

fun

ctio

n m

uta

tio

ns

in m

iner

alo

cort

ico

id

rece

pto

r

Part

ial l

oss

of f

un

ctio

n o

f th

e m

iner

alo

cort

ico

id

rece

pto

r, im

pai

rin

g sa

lt r

eab

sorp

tio

n;

imp

rove

men

t wit

h a

ge a

nd

a h

igh

-sal

t die

t

Dec

reas

ed

(co

nti

nu

ed)

K10373_C008.indd 269 10/8/2010 6:14:46 PM

270 ◾ Nanosensors: Theory and ApplicationsTa

ble

8.1

(con

tinu

ed)

Sing

le G

ene

Dis

ease

s Th

at E

leva

te o

r Lo

wer

Blo

od P

ress

ure

Dis

ease

Mu

tati

on

Mo

lecu

lar M

ech

anis

mEf

fect

on

Blo

od

Pr

essu

re

Lid

dle

’s s

ynd

rom

eM

uta

tio

ns

in th

e EN

aC β

or

γ su

bu

nit

Del

etio

n o

f th

e C

-ter

min

al d

om

ain

of E

NaC

, re

sult

ing

in in

crea

sed

EN

aC a

ctiv

ity

Incr

ease

d

Pseu

do

-h

ypo

ald

ost

ero

nis

m

typ

e I (

auto

som

al

rece

ssiv

e)

Loss

-of-

fun

ctio

n m

uta

tio

ns

in E

NaC

su

bu

nit

sIm

pai

rmen

t of E

NaC

su

bu

nit

s, w

hic

h is

no

t am

elio

rate

d b

y ac

tiva

tio

n o

f th

e m

iner

alo

cort

ico

id r

ecep

tor

by

ald

ost

ero

ne;

no

im

pro

vem

ent w

ith

age

; mas

sive

sal

t su

pp

lem

enta

tio

n r

equ

ired

Dec

reas

ed

Git

elm

an’s

syn

dro

me

Loss

-of-

fun

ctio

n m

uta

tio

ns

in th

e so

diu

m–c

hlo

rid

e co

tran

spo

rter

of t

he

dis

tal

con

volu

ted

tub

ule

Salt

was

tin

g fr

om

the

dis

tal c

on

volu

ted

tub

ule

, le

adin

g to

act

ivat

ion

of t

he

ren

in–a

ngi

ote

nsi

n

syst

em; s

ub

seq

uen

t act

ivat

ion

of t

he

min

eral

oco

rtic

oid

rec

epto

r in

crea

ses

ENaC

ac

tivi

ty, p

rese

rvin

g sa

lt h

om

eost

asis

No

rmal

or

dec

reas

ed

Bar

tter

’s s

ynd

rom

eLo

ss-o

f-fu

nct

ion

mu

tati

on

s in

gen

es r

equ

ired

for

salt

re

abso

rpti

on

in th

e th

ick

asce

nd

ing

loo

p o

f Hen

le

Salt

was

tin

g in

the

thic

k as

cen

din

g lo

op

of

Hen

le le

ads

to a

ctiv

atio

n o

f th

e re

nin

–an

gio

ten

sin

sys

tem

an

d th

e m

iner

alo

cort

ico

id

rece

pto

r, in

crea

sed

EN

aC a

ctiv

ity,

an

d r

elat

ive

salt

ho

meo

stas

is

No

rmal

or

dec

reas

ed

Sou

rce:

Rep

rin

ted

fro

m N

abel

, E.

G.,

New

En

glan

d J

ou

rnal

of

Med

icin

e, 3

49,

60,

2003

. W

ith

per

mis

sio

n.

© 2

003

Mas

sach

use

tts

Med

ical

So

ciet

y. A

ll R

igh

ts R

eser

ved

.

No

te:

ENaC

den

ote

s ep

ith

elia

l so

diu

m c

han

nel

.

K10373_C008.indd 270 10/8/2010 6:14:47 PM


cases,maynotidentifythereasonorprovidediagnosisbasedonthatdata,alone.Convergenceofdiagnostictoolsisnecessarytofurtheranydecisiononsuchdiag-nosis.Currenttoolsandtestsmaygraduallyundergochangeswiththeemergenceofpersonalizedhealthcare,madepossiblebysequencingofthehumangenome[92]anddevelopmentofgenomics-basedtools[93](seeSection8.3.4).

The illustration inFigure8.5hasyet anotherdimension that isparticularlyimportantinglobalhealthcare,especiallyforemerginganddevelopingeconomieswherethenumberofandaccesstomedicalexpertsarelimited.Thegoalistosup-port localdataanalysiseitherwithinformationorconnectivitythatmayenableaccesstoexperts,opinions,andresources.Variouseffortstodeliverinformation[94]throughICT[95]havebeenpioneered[96].However,thereisroomforfur-theradvanceswithdiffusionofbroadbandandotherhigh-speednetworkstothefarcornersoftheworld.Itmaytransformthevisionwhereapatientinaremotevil-lageclinicinMalawimayhaveaccesstoanelectrocardiograph(ECG)orlow-costRingSensorandcantransmitthatdatathroughastandardnetworkorinnovative3Gsystem[97]thatoffersmobilephoneserviceevenfromairplanes,duringflight[98].InthevillageinMalawi,anursepractitionermaybetheonlymedicalprofes-sionalintheclinicandmaybeunabletodeciphertheECGandhenceincapableof suggestingmedication.It ishere that thevalueof thetransmittedECGdatabecomesobvious.Throughaconsultancynetwork,ontheotherendoftheworld,acardiologist[99]mayreviewthedataanddiagnosecardiacarrhythmiaduetorepolarizationabnormality(clinicaleffect)causingLong-QTsyndrome[100].

Thissimpleexampleandothertypesofanalysis,which,inadditiontoconnec-tivity,mayalsorequirecomputationalpower,forexample,analysisofbrainactiv-itydatafrommagnetoencephalography(MEG),mayimmenselybenefitfromgridcomputing [101]. Advances in microfabrication of atomic magnetometers couldenablethedevelopmentofprecisionmagneticresonanceimaging(MRI)systemsforself-monitoring,inanylocation[102].

Thestrikingbenefitsthatmayemergefromthetrinityofgridinfrastructure,remoteMRImonitoring,andintelligentorexpertdataanalysis(AGRMIA)maybeappreciatedinviewofthefactthatmentalhealthanomaliesoftendisplaysignsthatmayresistdiagnosisduetolackofadequateexpressionofsymptoms.Individualsmayevenfailtorecognizethatsomethingisamissbecausetherateofincrementintheexpressionofsomementalhealthconditionsmaybeinfinitesimalandoversev-eralyearsordecades.Itmaynotbeuncommonthatindividualsmayevenadapttothesechangesas“agerelevant”ratherthandifferentiatingthemaspotentialsymp-tomsofadiseaseanddemandingmedicalexplorationortreatment.PersonalMRIdevicescoupledwithmicrofabricatedMSmaycreatetheremotems-MRIpersonalmonitoringdevicethatcouldworkasanoninvasivewearablewirelesssensorthatcanbeplacedontheheadtofitasaswimmingcap.Thisdevelopmentshallunleasha new horizon in “being digital” [103] in personalized medicine. In particular,remotemonitoringusingms-MRIsensorsmaybeinstrumentalforearlydetectionofbiochemicalchangesinthebrain,eithersporadicorduetoaging.

K10373_C008.indd 271 10/8/2010 6:14:47 PM


The immeasurable value of ms-MRI remote sensors may be best illustratedbyAlzheimer’sdisease. It is a conditionwhere the activityof the choline acetyltransferase(CAT)enzyme,responsibleforthesynthesisofacetylcholine,showsa60%–90%decrease[104].Acetylcholine isakeyneurotransmitterandamarkerforcholinergicneurons.Theabilityofms-MRItodetectandprofile(usingMS)biologicalandbiochemicalmolecules isthedrivingtechnologytodeterminetheshape and concentration of acetylcholine molecules and hence by extrapolationdeterminetheactivityofCAT.Thebiochemicalidentificationofmolecules(andinfutureidentifydifferencesinthestructureorshapeofthemolecules)iscriticalinAlzheimer’sdiseaseandms-MRIisonepromisingtoolthatisamenabletoworkasawirelesssensorforremotemonitoring.Thechoiceforms-MRIoverconventionalMRIisbasedonthefactthatconventionalMRIonlyidentifiesphysicalstructures.RecentdevelopmentsinfunctionalimagingusingMRIhavecreatedthefunctional-MRI(fMRI)thatcanidentifytherateofbloodflowwithinaphysicalstructureorareainthebrain.Hence,fMRImaybeusefultomonitorlearningdisabilitieswhereexternalstimulimayfailtoactivatecertainregionsofthebrain,suggestingabnormalities. However, in Alzheimer’s disease, the biochemical loss of enzymeactivityofcholineacetyltransferaseoccurs inthecerebralcortex,hippocampus,andrelatedareas,butcell countsof theneocortexandhippocampusofpatientswithAlzheimer’sdiseasedidnotrevealmajorreductionsinnumbersofcholinergicneuronswhencomparedwithage-matchedcontrols.Thus,forthepurposeofearlydetection,individualsdevelopingAlzheimer’sdiseasemayappear“normal”bycon-ventionalMRIanalysissincethephysicalstructureofthepotentiallyaffectedareasofthebrainremainsunchangedasfarasthenumbersofneuronsareconcerned.TheformationofplaquesinthebrainofpatientsaffectedbyAlzheimer’sdiseasemaybedetectedbyMRI.Remotemonitoringusingms-MRIwirelesssensorsmaybealsoapplicabletootherconditionsrelatedtochangesinneurotransmitter-relatedproteinsandmoleculesinthebrain,forexample,Parkinson’sdisease,Huntington’sdisease,andsomeformsofdementia.

Theadoptionofthesedevelopmentsinpersonalremotemonitoringofmentalhealthcoupledwiththeabilityofindividualstoobtainanexpertopinionbytrans-mittingthedataaswellaslearnabouttheimplication,ofthechangesrecordedovertime,mayhelpdetermineamedicalmanagementplanthatmayimprovetheindi-vidual’squalityoflife.Theanalysisofdatamayrequirecomputationalresourcesthatmaybeunavailableinmanylocations,eveninaffluentnations.Theabilityofthetransmitteddatavia the localnetworktoaccess“medicalgrid”-basedexpertservicesoffersimmensebenefits.Theaccesstotheseresourcesthroughthenetworkandmedicalgridservices,evenfromthedevelopingnations[105],canreshapethefabricofglobalmentalhealth.

InadditiontoMRI,gridcomputinghasthepotentialtoaddremarkablevaluetootherformsofremotebiomedicalimagingsystemsaswellasbio-telematics[106]sincecurrentoperations[107]arelimitedtogroups[108]thatengageinstand-alonepoint-to-pointsystems[109]withoutthebenefitofaplatformtoaggregatemedical

K10373_C008.indd 272 10/8/2010 6:14:47 PM


gridtypeservices.Bringingtogetherthisplatformmaybesimilarinefforttocreat-ingtheGRMandisexpectedtobeapartoftheGRM(Figure8.2)thatmayrunonagridinfrastructure.Severalbiomedicalimagingdatabasesareinexistence,andtakentogether,theymayformabiomedicalimagingplatform(BIP)tetheredtoaproposed“bMDs”servicesinfrastructure[110]throughanaccessibleopenformatwhere images (data) may be uploaded from anywhere in the world and viewed(analyzed)byexpert(s)whowillshareobservationsand/ordelivertheirinterpreta-tionordiagnosisdirectly to thepatientor thehealthcare serviceprovider.Toolsforsimulationandvisualizationareimportantandsignificantadvances[111]canberesourced.Agentsembeddedinthearchitecturemaymonitor,devicetodevice(D2D), machine to machine (M2M), and device to system (D2S) or vice versa(S2D),medicaldatatoensuredatasecurityandprivacyissues.

Progress in the AI vision of autonomic computing may gradually transformBIP,eitherindependentlyorincombinationwithanAGRMIAtypeinfrastructure.Inclusionofembeddedintelligencemayprovideopinionsorrecommendationsordiagnosisorreferrals(exceptionmanagement)withoutactivehumanintervention.ThelattershouldbewelcomenewstotheagingpopulationinEuropeandJapanwhowishtoremainindependentandliveintheirownhomesratherthaninlong-termhealthcarecommunitiesthatcandrainnationalhealthcareresources.Nationsmaybeprudenttoexplorems-MRItypehightechnologymedicalpracticesandfindnewwaystothinkaboutdiseases[112]withlong-termimpactandchallengethemedicalgovernance[113]toreducecostbyinvestinginandacceleratingtheconvergenceofmedicalknowledgeandengineeringtechnologies.TheconvergenceproposedinbMDsshallincludeadvancessuchasms-MRIremotemonitoringplat-formsandmaybeapartoffuturehealthcaretapestry.

Thus, grid computing is an enabling technology for healthcare service con-nectivityinmuchthesamewaythatgridcanfacilitatebusinessservices[114].Inhealthcare,asinbusiness,gridcomputingmayprovideaccesstoon-demandcom-putationalresources(thatareinadifferentlocation)forreal-timedataprocessingandanalysis,throughgrid-basedtools,suchasGlobus[115].

8.3.4 Innovation in Wireless Remote Monitoring and the Emergence of Nano-Butlers

Nano-butlersisafacetioustermbutisexpectedtoconveyan“image”tosuggestthatnano(small)-toolsandtechnologiesactas“smallbutlers”servingthedemandsofhealthcare.(Their“fees”arealsosmall;hence,theyworkfor“micro”payments.)The“tongue-in-cheek”imageisexpectedtocreateanawarenessthatthedetectionofbiologicalmoleculesincludingproteinsorpeptidesatnanolevelsmaybecriticalfortheidentificationofbiomarkersthatmaybeassociatedwiththeriskofadisease.When the investment to develop nano-detection is recovered (return on invest-ment)fromthesavingsfromreducedacute-careresponses, ifmaybe,theactualcostofnano(small)-detectionwillbesmallenoughtoenableglobaldiffusionofthe

K10373_C008.indd 273 10/8/2010 6:14:47 PM


tools,whichcanbesustainedbymicro(small)payments,mimickingtheconceptof micro-finance that has gained global acclaim [116] for alleviating poverty insomepartsofthedevelopingworld.Thisapproach,duetotheinclusionoftheterm“nano,”mayalsodrawsomeunfoundedattention[117].

Early risk identification,prior todetectable symptoms,offers thepotential todevelopamanagementplantocontainthediseaseorevenstopitfrompresentinganysymptoms.Thisapproachreducestheacute-careresponsethatmaybenecessaryiftheconditionwasnotdetectedandleftunattended.A&Eresponsesandacute-carearefarmoreexpensiveandoftenincreasemorbidityandmortalityiftheresponsefailstobeadministeredinnearreal-time.Forexample,fromthetimeofonsetofacardiacattack,thereisonlya30minwindowforsuccessfuladministrationoftissueplasminogenactivator(TPA)todissolveclot(s)ifapatientwithCVDorstrokeisthevictimofthrombosis(bloodclot).Itislikelythatmorethan30minmayelapsebetweenrecognizingthatapersonishavinga“heartattack”andthearrivalofA&Eservicesatthelocation,assumingnotonlythattheparamedicswillcorrectlydiag-nosethereason(clot)butalsothattheywillhaveaninventoryofTPA-typedrugsintheirmobileunit(ambulance).ItisalsoassumedthatthepersoniscoordinatedandcoherentenoughtocalltoA&Eservicesiftheindividualisaloneinthelocation.

Preventionthroughpreventativehealthcaremayrequire,intheabovescenario,the convergence of wireless sensors with remote monitoring technologies anddata-drivenanalyticsbasedonbiomedicalresearchknowledgebases.Advancesinunderstanding the basic physiological, biochemical, and molecular relationships(Figure 8.6) that contribute to heart disease [118] offer hope for rigorous earlydetectionmechanisms.Thereisampleevidencebothfrombasicbiologicalsciences[119]andclinical research [120] that earlywarning signsofCVDareamenableto identification from research on biomarkers. Several biomarkers for CVD arealready in the market [121], but the commercial “kit” approach is far from theinnovativepotentialofnextgenerationdiagnostics[122].

Innovationindetectionisbasedonthegenerallyapplicableprinciplethatphysi-ologicalsystemsrespondtothresholds.Inotherwords,few,ifany,reactionsoccurinthehumanbodyorfetuswithoutacriticalmassorconcentrationofmolecules.Ifallowedtoreachthe“threshold”only,thenthe“rogue”moleculesmaytriggera cascade of events,whichmay, eventually, over time, present itself as a detect-ablesymptom.Hence,molecularidentificationofthebiomarkersandroguemol-ecules isvital fornoninvasivedetection.Tools fordetection requireconvergenceoftheknowledgefromidentificationofmoleculesfrombiomedicalresearchwithengineering-baseddetectiontechnologiestodeterminethenumberandconcentra-tionofthemolecules,beginningatthesinglemolecule[123]leveloratthepico(10–12)level,butmostreliablyatthenanolevel.Amedicalmanagementplanortreatmentmustexisttopreventtheconcentrationofthemoleculesfromreachingthethreshold,whereitmaycommencethecascadeofeventsleadingtoadiseaseorsymptomorprecipitatingaheartattack.Incaseoffetaldiagnosis,themanagementissuesmaybecomplex,buttheearlydetectionofsporadicorgeneticdiseasesofthe

K10373_C008.indd 274 10/8/2010 6:14:47 PM


unbornchildmayofferscopeformedicalintervention.Foolproofidentificationofspecificbiomarkersformulti-factorialdiseasessuchasCVDiscomplex.Withoutreliablespecificity,thenextgenerationnano-diagnostictoolsmayofferlessvalue.Therefore,thehealthcareindustrymustremainvigilanttocombineadvancesinonefieldwithanotherthroughparallelinvestmentsbothinmedicineandengineering.

Toillustrate, letmerevisit thecaseofbloodpressure(BP)measurementandwhatdiagnosticinformationtheBPdatamayprovideiftheBPiselevated,lower,ornormal,comparedtothestandardreference.Inshort,theBPdata,alone,providelittlediagnosticvalueunlesstheyareanalyzedinconjunctionwithexistingcasehistoryorotherdata.OnereasonforthisissummarizedinTable8.1.IfadefinitivediagnosisoftheBPdataissought,insomepatients,itwillbenecessarytoidentifywhichparticulargeneisaffected.

Sequencingof thehumangenomeandadvances in search tools andgenomictechnologies[124]makesitpossibletoextracttheDNAsequencefromtheHumanGenomeDatabase[125]ofthegenesimplicatedinadisease(e.g.,genesthatmayelevate or lower blood pressure). Based on the DNA sequence, anti-sense RNAmaybeusedtodeterminegeneexpressionprofile.ByextrapolationfromtheDNA

LDL receptor

Familial hypercholesterolemia

Autosomal recessive hypercholesterolemia

Familial ligand-defective apolipoprotein B-100

Apolipoprotein B-100

PhospholipidsCholesterol

ester core

Cholesterol pool

Liver

LDL

(–)

Plasma

Fat

Artery

Atherosclerotic lesion

Triglycerides

TriglyceridesVery-low density lipoproteins

Sitosterolemia

Cholesterol absorption

Intestine

Dietary fats

Bile

ABC G5ABC G8

Muscle

Figure 8.6 Components of cholesterol synthesis and excretion. (Reprinted from Nabel, E.G., N. Engl. J. Med., 349, 60, 2003. With permission. © 2003 Massachusetts Medical Society. All Rights Reserved.)

K10373_C008.indd 275 10/8/2010 6:14:50 PM


sequence,protein-basedpeptidefragmentscanbesynthesizedforuse innoninva-siveproteomics-basedmicroarrays,usingthelab-on-a-chipwirelesssensor,todetectexpression levelsofoneormoreproteins and/or theirmutantvariations, inbodyfluids, thatmaybe involved in the etiologyof the conditionunder investigation.Basedonthegeneandproteinexpressionprofile,insomecasesandsomediseases,thepromiseofgenetherapymayberealizedby“silencing”harmfulcandidategenesusingRNAi[126]andsnRNAtechniquesthatarealreadyunderintensecommercialexplorationandhavereportedsomedegreeofsuccess[127]for futuretherapeuticapplications.

This scenario, startingwith apossible gene, followedby expressionprofilingdataremotelymonitoredbyawirelesssensorandpotentialforselectivesilencingtherapyofthedisease,isapartoftheevolutionlikelytochartthefutureofindi-vidualized medicine and personalized healthcare. The value and impact of thisapproachmaynotqualifyasa“disruptive innovation”[128]butmayreflect thesystemiclessonsfromtheageofintroductionoftheelectricdynamo,attheturnofthetwentiethcentury[129],thewisdomofwhichmayhavebeenignoredbyotheremergingtechnologies[130].

Thepragmaticcredibilityofthisvisiongarnerssupportbothfromcurrentprac-tices,albeitinpart,andinformationbasedonrecentresearchthatclearlypointstotheneedforthisvision.Currentpracticesalreadyuseoneormoreofthestepsoutlinedinthisstrategy,forexample,theuseofgenomic[131]andproteomicbio-markersinthediagnosisofCVD[132]andsomeformsofcancer[133].Butevenmoreimportantsupportforthisvisiondrawsonseminalmedicalresearch[134]thathasidentifiedonesinglehumangene,forlow-densitylipoprotein(LDL)recep-tor-relatedprotein6(LRP6),thatisinvolvedintheetiologyofaspecifictypeofCVD,referredtoascoronaryarterydisease(CAD).Itisaleadingcauseofdeathworldwide,andearlydetectionisessentialtosavelives.

CADiscommonlycausedbyaconstellationofriskfactorscalledtheMetabolicSyndrome,thesymptomsofwhichincludehyperlipidemia,hypertension,diabe-tes,andinaddition,osteoporosis.AnalysisofLRP6,thegeneidentifiedasrespon-sible for CAD, has uncovered a single nucleotide substitution that changes thewild-type (normal) cytidine base to thymidine. This single nucleotide missensemutation, located in theprotein-coding region (exon9)of thehumangene forLDLLRP6,causesasingleaminoacidsubstitution,whichinsertstheaminoacidcysteinetoreplacethenormalcounterpart,arginine,atcodon611(R611C).TheimportanceofthisfindingandthefundamentalsignificanceoftheLRP6geneinhumanevolutionarefurtherhighlightedbytheextremelyhighdegreeofconserva-tionoftheproteinsequencefromhumanstoamphibians,suchasfrogs(Xenopus laevis).Amongthespeciessurveyed,theaminoacidarginine(R)isconservedfromfrogstohumans(Table8.2).Substitutionofargininewithcysteine(R611C)cre-ateshavocandresultsinCADandaccompanyingdiabetes,hyperlipidemia,hyper-tension,andosteoporosis.

K10373_C008.indd 276 10/8/2010 6:14:50 PM


Tabl

e 8.

2 C

onse

rved

Am

ino

Aci

d A

rgin

ine

(R)

Is S

ubst

itut

ed in

LR

P6 G

ene

and

Cau

ses

Hea

rt D

isea

se

Hu

man

CL

YR

PQ

GL

RC

AC

PI

GF

EL

Ch

imp

CL

YR

PQ

GL

RC

AC

PI

GF

EL

Mo

nke

yC

LY

RP

QG

LR

CA

CP

IG

FE

L

Mo

use

CL

YR

PQ

GL

RC

AC

PI

GF

EL

Rat

CL

YR

PQ

GL

RC

AC

PI

GF

EL

Do

gC

LY

RP

QG

LR

CA

CP

IG

FE

L

Co

wC

LY

RP

QG

LR

CA

CP

IG

FE

L

Op

po

sum

CL

YF

PQ

GL

RC

AC

PI

GF

EL

Ch

icke

nC

LY

RP

QG

LR

CA

CP

IG

LE

L

Fro

gC

LY

FP

QG

PR

CA

CP

IG

LE

L

K10373_C008.indd 277 10/8/2010 6:14:50 PM


GenomictechnologiescanhelpidentifyR611Cmutationinat-riskpopulationseveninthefetal statesincethisLRP6is transmittedasanautosomaldominanttrait.Lifelongmanagementplanforinheritedgeneticdiseases,forexample,phe-nylketonuria,iscommon.IndividualshomozygousforR611C,ifdiagnosedearly,may follow a recommended lifestyle thatmay enable them to enjoynormal lifeexpectancy.TheevolutionaryconservationofLRP6geneanditsproteinproductindicatesafundamentalroleofthisproteininphysiology.Indeed,LRP6isinvolvedincellularsignalingpathways,disruptionofwhichleadstoaplethoraofproblems.Proteinsandothermolecules, referredtoas transcriptionfactors [135],areoftenconserved and fundamental to gene expression [136] from bacteria to humans.Transcription factors canaltergeneexpressionpositivelyandnegatively [137]ormayserveassecondaryortertiarytargets[138].Earlydetectionofnonfatalmuta-tionsintranscriptionfactors iswarrantedbecausetheymayalsocauseprofoundphysiological disturbances and present multiple symptoms, related in scope tomutationinLRP6.

Whilethecaseforearlydetectionofgeneticdiseasesneedslittleemphasis,theneedforearlydetectionofthebulkofsporadiccasessuchastypeIIdiabetesmel-litusandseveralotherdiseasestatesdeservesemphasis.Globalizationhasalsocre-atedaneedforremotemonitoring.Tomakeglobalizationworkbetterfortheworldeconomy[139] it is imperativetocontaininfectiousdiseasesbydeterminingtheriskandat-riskfactorsattheoriginratherthanatanimmigrationcheck-pointofacountry.ThelightningspreadofSARS,fromitsorigininHongKongtoToronto,inafewdays,highlightstheneedforremotemonitoringtoolsandthevulnerabilityofcurrenthealthcaresystem(inmostnations)thatisill-equippedforglobalchal-lengesandmayactuallyaidanepidemicorpandemic.

Epidemics,however,arenolongerlimitedtoonlyinfectiousdiseases.TypeIIdiabetesmayreachnearlyepidemicproportionsinmanynations.Itisalarminginsomecountries,where,despiteasmallpopulation[140]ahighnumberofsporadiccasesofdiabetesaredocumented.Inaddition,anequallyhighnumberofprojectedat-riskpopulationareacknowledged,butthelatterestimatesexcludethesegmentofpopulationprojectedtobedefinedasclinicallyobese.Thatcanpotentiallyincreasetheat-risknumbersforsporadictypeIIdiabetesbutremainsunaccountedbythesystem.Individualswithotheranomaliesmayalsohavediabetes,aspointedoutincourseofthediscussiononLRP6whereheartdiseaseanddiabetescanoccursimultaneously.Recentevidencesuggeststhatinindividualswithoutanygeneticpredisposition,theremaybeadirecteffectofelevatedcholesterollevelonreducinginsulinproductionbytheβ(beta)cellsinthepancreas[141].Thisincreasestheriskofdiabetesforobeseaswellasfornon-obeseindividualswhohaveelevatedlevelsofcholesterol,withoutgeneticpredisposition.

Diagnosed diabetics often monitor their blood glucose levels using over thecounterkitsbutitsuseinpreventativehealthcareremainsdubious.Fordiabetics,thefrequencyoftestingusingkitsisweeklyordailybutexpertinterpretationand

K10373_C008.indd 278 10/8/2010 6:14:50 PM


advicemaybefarinbetween.Letusassumethataweeklyoutpatientvisittotheclinic for blood glucose test and insulin therapy costs thehealthcare service anaverageof$25indirectcostandcoststheeconomyanother$25inindirectcosts,suchasthecostoftimespentbypatient,costoftravelanddecreaseinproductivityduetotimetakenoffworkbypatient.Ifonly1%ofthepopulationrequireweeklyattention(seriousdiabetics),thenforahypotheticalnationwithapopulationof5million,therewillbe50,000diabeticsrequiringthisattention.For50visitsayearat$50avisit,theattentionto50,000diabeticsforsimplemonitoringofbloodglucoseandadministrationofinsulin,willcostthenation$125millionperyear.Thefocusonthosewhoneeditthemostaggravatestheunattendedconditionsinotherdiabeticsandat-riskpopulation,drivingthemtoseekacute-careservicesorcatapultingthemtothe$50category.However,afargreaterconcernistheneedfor inpatient services if some of the diabetic patients need hospitalization. Canthenationalhealthcaresystemrespondadequately[142]ifonly1%ofthe50,000diabetics(thatis,1%ofthe1%documenteddiabeticsinthepopulation)needhos-pitalbeds?Forexample,inIreland,overall,thereare2.9bedsavailableper1000people[143].

Preventativeremotemonitoringcanalterthisviciouscycleofcrisis,reducecost,andimproveactualcare.

Apart from individuals who are obese or juvenile diabetics or those withhistoryofgeneticpredispositiontoearlyonsetdiabetes,thedefinitionof“earlystage” inmonitoring is rathervague for sporadicdiabetes.There is little scien-tificrationaleeithertoincludeortoexcludeyoungadultsandindividualswithdynamic vivacity in the prime of their life and in their 40s or even younger.Hence,effectivemonitoringofbloodglucosethat isnotshunnedbytheother-wisehealthypopulationmayrequirealifestyleapproachthatoffersaproductandservicethatareeasytouse,oflowrisk,oflowmaintenance,sociallyacceptable,ofrobustvalue,safe,andmedicallyeffective.Sincetheadoptionofthisdevicemaybevoluntary,itmaybelessattractiveifthemonitorisavisiblewearable[144]ora“thing”thatanindividualmust“remember”todo.Thus,nanotechnology-based[145]monitoringtoolsmayfunctionas“always-on”wirelessnano-sensorswhichremainunder theepidermis. Itmaymakebloodglucosemonitoringofgeneralpopulationanattractivemodusoperandiforearlyriskidentificationandpreven-tionofdiabetesaswellasassociatedmorbidity,suchasdiabeticglaucoma,whichcanleadtopartialorcompleteblindnessinseverediabeticsorjuvenilediabetics,ifleftuntreated.

Earlydetectionofdiabetes as a functionofmonitoringbloodglucosecon-centrationbenefitsfromaplethoraofglucosesensorsdevelopedoverthepast25years,butchallengesstillexist.Akeyadvantageinthedevelopmentofaminia-turizedornanoscaledevicethatcanquicklyandreliablymonitorglucoseinvivo,is basedon the fact that the level of bloodglucosedetectiondoesnot requirenano-leveldetection.Thebenefitofanano-deviceornano-sensoristomakethe

K10373_C008.indd 279 10/8/2010 6:14:50 PM


monitoringtoolvirtuallyunobtrusivetotheuser.Thenormalclinicalrangeforblood glucose is in the millimolar (mM) range between 3.5 and 6.1mM, butabnormalglucoselevelsmayreach20mM.Thisconcentrationrangecanbeeas-ilymonitoredusingelectrochemicalreactions.Whatiscriticalfordiabetes isatoolthatdoesnotdeterearly-stagefrequentvigilanceaboutthechangesinbloodglucoselevels,alsomeasuredinmilligramsperdeciliterofblood.Alterationsmaysignaltheabilityorinabilitytomaintainthestandardequilibriumconcentrationofglucose(120mg/dLor3.5–6.1mM).Forexample, ifthebloodglucosecon-centrationinanindividualtakeslongertoreturntonormalaftermeals,thenitmaysignalgerminatingproblemswithglucoseclearanceand/ortoleranceintheindividual.

Theinnovationrequiredtodevelopawirelessbloodglucosenano-sensorthatiscapableofsubcutaneousmonitoringofbloodglucoseandtransmissionofthedatatoawirelessnodemaybesimpleandathand.Thecorecomponentsareanano-sensor[146]capableofdetectingbloodglucoseconcentrationinvivoandanano-radio[147]thatcantransmitthedata.ThecombinationisillustratedinFigure8.7.

RE (Ag)

Glucose, O2Glucose oxidase

Conducting polymer

Gold nanoelectrode20–60 nm

A

B potentostat

Nanotube radioBlood glucose nanosensor

Speaker

Radio signal

Positive electrode

+

+

–

–

Negative electrodeCarbon

nanotube

Battery

Eg

Veg

CE (Pt)

H2O2

Figure 8.7 Blood glucose monitoring and wireless data transmission 24.7.365. (Adapted from Forzani, E.S. et al., Nano Lett., 4, 1785, 2007; Jensen, K. et al., Nano Lett.,7, 3508, 2007. With permission. © 2007 American Chemical Society.)

K10373_C008.indd 280 10/8/2010 6:14:52 PM


Theopenquestionspresentedbythisinnovativepotentialmaybedividedintotwo broad categories: data acquisition and data transmission. The issues are asfollows:

1.Glucose sensor data transmission by the nanotube radio illustrated inFigure 8.7 and other similar [148] devices is not proven because they areconstructed as receivers,notdata transmitters.But, in general, single-wallnanotubes(SWNT)arelikesingle-modefiberforelectrons(Figure8.7)andhencehavedatapropertiesthatweremadetoactasareceiver(Figure8.7)forthenano-radiobutmaybealteredtotransmittheacquireddatafromtheglucosesensor.

2.Functionalco-fabricationorsimplyco-locatingor“housing”glucosenano-sensoranddatatransmitteronanonallergenicmatrixorplatformsuitableforuseasasubcutaneousimplant.

3.Optimizingsignal-to-noiseratio. 4.Interferenceminimizednano-communicationlinktowirelesssensornode. 5.Physicallocationsforsafesubcutaneousinsertionpercustomerpreference. 6.Procedureforsafeextractionofsensordevicewithminimaldiscomfort. 7.Foolproofimmobilizationofimplant. 8.Containmentofdegradationorbreakageofcomponentswithinthehousing

ofthenano-device. 9.Customer’sabilityto“forget”aboutsensorimplant. 10.Customercontrol(agent-basedwebtool)overfunctionofthedevice. 11.Customer control over data transmission or modulating the frequency of

monitoring. 12.Lowmaintenanceofsensorandtransmitter. 13.Batterylifeoftransmitter. 14.Exploreuseofelectrolytegradientofthebodyorenergyfrommovementto

powerdatatransmission.

The responses to the open questions are beyond the scope of this chapter, butthedifferentmechanismsofdetectionbysensorsareimportanttoreview,briefly,becausewirelessnano-communicationmustbe an integralpartof the sensororsensorcombinationinordertoreliablytransmitthedataoutsidethebody.

TheglucosesensorinFigure8.7detectsglucosebasedonprinciplesofelectro-chemistry.Theassumptionisthatthesensorwillperforminvivoaswellasithasperformedinbodyfluidstestedinvitro.Specificdetectionofglucoseismediatedbyglucoseoxidase,GOx,anenzyme(hollowcircleswithpinkborders,Figure8.7)that is immobilizedontoPANI-PAA, a conductingpolymer (green area,Figure8.7)madeupofpolyaniline (PANI)polymerizedwithPAA (poly{acrylic acid}).Uponexposuretoglucose,GOx,withthehelpofanaturalcoenzyme,flavinade-ninedinucleotide(FAD),catalyzestheoxidationofglucosetogluconolactoneandbecomesreduced,{GOx(FADH2)},Step1.ThereducedGOxform,{GOx(FADH2)},

K10373_C008.indd 281 10/8/2010 6:14:52 PM


isregeneratedviareoxidizationbyoxygen(O2)insolutiontoGOx(FAD)andpro-duceshydrogenperoxide(H2O2),Step2.Polyaniline,whichexistsinitsreduced(red)state(PANIred),isoxidized(ox)byH2O2toPANIoxandtriggersanincreaseinpolyanilineconductivity(Step3)duetothesensitivedependenceofpolyanilineconductivityonitsredoxstate.Thischangeofconductivityisdata,indicatingglu-cosedetection.

Step1:Glucose+GOx(FAD)→gluconolactone+GOx(FADH2)Step2:GOx(FADH2)+O2→GOx(FAD)+H2O2Step3:H2O2+PANIred→H2O+PANIox(increaseinconductivity)

Other types of construction use carbon nanotubes (CNT) but may stilluse the electrochemical principles for detection of glucose on a CNT scaffold(insteadofPANI-PAA,asshowninFigure8.7).Nano-wires[149]representonesuchtypeofconstruction.Nano-wire-basedglucosebiosensors[150]useCNTnano-electrodeensembles(NEE)forselectivedetectionofglucosebasedonthehighelectro-catalyticeffectand fastelectron-transfer rateofCNTbutemploythesameelectrochemicalmechanismdescribedabove.GOx,glucoseoxidase,isimmobilized on CNT-NEE, instead of PANI-PAA, via carbodiimide chemis-trybyformingamidelinkagesbetweentheamine(NH2)residuesandcarbox-ylicacid(COOH)groupsoftheenzyme,GOx,covalentlylinkedtotheexposedtipsofsingleCNT,illustratedasperpendicularblackbarsinFigure8.8(center).NumbersofCNTonaCNT-NEEareinthemillions,witheachnano-electrodebeinglessthan100nmindiameter,thereby,increasingsensitivityofthesensor(by analogy, the speedof a processor, e.g., IntelPentium, is a functionof thenumberofmicroprocessorcircuitsetchedonthechip).Thecatalyticreductionofhydrogenperoxide liberated fromtheenzymatic reactionofglucoseoxidasecovalentlyimmobilizedontheCNT-NEE,inthepresenceofglucoseandoxy-gen,leadstotheselectivedetectionofglucose.CNTareexcellentelectrochemicaltransducers,andeachCNTservesasanano-electrodethatdetectsthechangeincurrent(conductivity)whenglucosereactswithGOx-linkedCNT(thecouplingdrives the specificity of glucose detection) in a CNT-NEE sensor. The sensoreffectivelyperformsaselectiveelectrochemicalanalysisofglucoseinthepresenceofinterferencefromcommonmolecules,forexample,acetaminophen{AA},uricacid {UA}, and ascorbic acid {AC}, shown inFigure8.8 (toppanel).Butmostimportant, the sensor is sensitive to increments of glucose. Detecting fluctua-tionsinconcentrationofbloodglucoseisthekeytoearlydetectionindiabetes(Figure 8.8,lowerpanelandinset).

It is relevant tonote thatnano-wire sensors thatdetect changes inchemicalpotential,accompanyingatargetoranalytebindingevent,suchasDNAorRNAhybridization,peptideinteractions,oroxidationreductioninelectrochemicalreac-tions,canactasafieldeffectgateuponthenano-wire,therebychangingitsconduc-tance.Thisissimilar,inprinciple,tohowafield-effecttransistor(FET)works[151].

K10373_C008.indd 282 10/8/2010 6:14:52 PM


Today,FETsarelowcostandinextensiveuseinenvironmentalandagriculturalmonitoring.Advancesintechnologyhavemadetheexpensiveelectronicmarvelofa“transistorradio”ofthe1950sonlysuitableforinfant’stoysmassmanufacturedinChinaandsoldindiscountchains.ThetransistorradiohasevolvedtodirtcheapFETsnowusedinanimalfarmstoalertownersthatthestenchfromtheammonia-richwastefromanimalexcretaintheholdingtanksneedsattention.Theexpensivetransistorofyesterdayisalow-cost(ornocost)componenttodaythatdeliverssig-nificantvalueforenvironmentalmonitoring.Itbenefitsthemeatindustryatsuchanegligiblecostthatithasonlyincreasedproductivityofthemeatindustryandconcomitantincreaseinglobalconsumption.TheevidenceforthelatterisgleanedfromthebeefandchickenconsumptiondatafromtheUnitedStatesandEUthatexceeded120kgperpersonperyear(330g/day)comparedto16kgperpersonperyear(44g/day)inChinaandIndia,combined[152].

–4.0

–3.5

Glucose speci�city

AA UA AC

HN-GOx

C=O C=O C=O C=O

HN-GOx HN-GOx HN-GOx

Curr

ent/1

e-8A

Curr

ent/1

e-8A

–3.0

–2.5

–2.0

–1.5

–1.0

–0.5

–0.5

–4.00 50 350 400 Time (s)

Time (s)

–3.5

–3.0

–2.5

–2.0

–1.5

–1.0

0

0

2 mM Glucose

Glucose concentration (mM)0

010 20

20

406080

30

Curr

ent

50 100 150 200 250 300 350 400

Figure 8.8 Nano-wire glucose sensor shows glucose specificity and sensitivity to concentration. (Adapted from Lin, Y. et al., Nano Lett., 4, 191, 2004. With permission. © 2004 American Chemical Society.)

K10373_C008.indd 283 10/8/2010 6:14:53 PM


KnowledgefromresearchandapplicationsofFETareavailablefromarchivesthatmaydatebacktotheinitialdiscoveryoftransistors[153],nearly70yearsago.Thedevelopmentofnano-wire,nano-sensors,andnano-communicationmayben-efitfromtheexperienceandwisdomoftheFETpioneers.

TheprincipleofFETindeedmaybecrucialfornano-communicationforwire-lesstransmissionofdatafromtheinvivosensortoasensorcommunicationlinkornodeoutsidethebodythatcanconnecttotheInternet.ThecharacteristicofaCNTtoactasfiberthatcantransportelectronsisunderscrutinyinordertousethematerial forhigh-speeddata transmission in a variety ofways that includesthe emerging field of plasmonics [154] that studies interactions between lightandnanoscaleparticlesandstructures.Remotemonitoringinvivothatmaytakeadvantageoflight-emittingluciferase[155]enzyme-linkedsensors,byimmobiliz-ingluciferase(onPANI-PAAmatrixor)preferablyonCNT-NEEtypescaffolds,mayfinditusefultoexploitthepotentialofnanoscaleantennaethatconvertslightinto broadband electrical signals capable of carrying approximately one milliontimesmoredatathanexistingsystems[156].

TheillustrationinFigure8.7issimpletounderstandandconveystheimageof the components necessary to drive the convergence of in vivo detection andtransmissionofdata.But,cautionisnecessarytoextrapolatetheapplicationofthecomponentsillustratedinFigure8.7.Althoughitmaybeidealforvisualizingtheconcept,thecomponentsillustratedarenotprovenorguaranteedtobethecom-binationofchoicethatcoulddrivethedevelopmentofaninvivonano-devicethatisequallyreliableasadetectiontoolandanano-communicationtool,forreasonsthatIshallexplaininthenextfewparagraphs.Butofcoursethatargumentholdsforanyinnovation.Weshallnotknowtheoutcomeunlessweattempttocreatethedeviceifthereiseven“justenough”reasonthattheinnovationmaybearfruit.Reasonably,oneissue intheglucosenano-junctionsensor is that itdoesnotusenano-materials,thatis,inthiscase,CNT,initsconstruction.Nano-wireglucosebiosensorsuseCNTasthenano-electrode,asshownintheCNT-NEEsensorillus-tratedinFigure8.8(center).Thismaybeanissueintermsoftheabilitytotransmitthedataoutofthebody.

If the nanotube radio receiver illustrated in Figure 8.7 can be modifiedordesignedasan invivo transmitterand if it can successfullydetect thedataemergingasthechangeinconductivitybetweenthetwostatesofPANI(PANIreducedversusPANIoxidized,asshowninStep3)fromthenon-CNTglucosenano-junctionsensorillustratedinFigure8.7,thenthenano-devicemaybepro-duced,bythecombinationillustratedinFigure8.7,forwirelessinvivoglucosemonitoring.

ThepreferenceforCNT-basedsensorslikeCNT-NEE(Figure8.8,center)islinkedtothecentralneedtotransmitthedatafromtheinvivosensorandthepotentialforusingindividualnanotubeswithinCNTnetworkstocarryinforma-tion. Innovation innano-communicationusingCNTcommunicationnetwork[157]islikelytobecomeacorecompetencynecessaryforthefutureofnano-device

K10373_C008.indd 284 10/8/2010 6:14:53 PM


useinhealthcare,ingeneral.Itisinthiscontextthattheprincipleoffieldeffecttransistors(FET)springsbackintoaction.CurrenttechnologyutilizesanentireCNT network as semiconducting material to construct a single FET. SeveralFETsarerequiredtobuildtraditionalorlegacynetworkequipment.Theresultisthattherearemanynanoscalenetworksembeddedwithineachdevice(FET)thatmightbeotherwisemoreeffectivelyutilizedforcommunication.In otherwords,theCNTnetworkitselfisthecommunicationmedia,andindividualCNTarethelinks.But,individualCNTandtubejunctions(formingnodes)donothavetheequivalentprocessingcapabilityofatraditionalnetworklinkandnetworknode.Tocompensateforthis,thesystemneedstoleveragelargenumbersofCNT.Theillustration inFigure8.8 (center) shows individualCNTlinked toGOx (blackbars),butmillionsofCNT(blackbars)makeupanensemble(CNT-NEE).Thelatterispreciselywhatisnecessaryforthesingle-wallCNTcommunicationnet-work(NanoCom)ofthefuture.However,NanoCommaynotfunctionaccordingtothetraditionalarchitectureofdatacommunicationlayers[158].ComparisonoflegacycommunicationandNanoComhighlightsthechangesnecessary,asshowninTable8.3.

At the bottom level (least sophisticated level), communication links may bebetweenhostsandroutersinacommunicationnetwork,ortheymaybeCNTover-lappingatpointsthatwillbeidentifiedasnodes.Anetworkfunctionsbychangingstate.Datamusteitherfloworbeswitchedorroutedthroughnodes.StatemaybeimplementedasaroutingtableonarouteroranelectromagnetfieldcontrollingtheresistancewithinaspecificareaofaCNTnetwork.Finally,amechanismneedstobeinplacetocontrolstate(ascendinglevelofsophistication,Table8.3),beitaroutingalgorithmorFETgatevoltagesappliedtoaCNTnetwork.Thetraditionalnetworking protocol stack is inverted in this approach because, rather than thenetworklayerbeinglogicallypositionedabovethephysicalandlinklayers,asinthe

Table 8.3 NanoCom Requires Modification of Current Networking Concepts

Basic Network Components

Traditional Networking

Nanoscale Networking

Increase in conceptual sophistication

Protocol Processors Gate control

State Node memory Semiconducting tube resistance

Network Links Nanotubes

Source: Bush, S.Y. and Li, Y., Nano-Communications: A New Field? An Exploration into a Carbon Nanotube Communication Network, Technical Information Series, GRC066, GE Global Research Center, Niskayuna, NY, 2006. © 2006 Inderscience.

K10373_C008.indd 285 10/8/2010 6:14:53 PM


standardOSImodel[159],theCNTnetworkandroutingofinformationareanintegralpartofthephysicallayer.

Data transmission in a CNT network occurs via modulated current flow(changes in conductance) through the CNT network guided towards specificnano-destination addresses.The addresses identify spatiallydistinct areas of theCNTnetworkthatmaybemadeupofnanosensorarrays.SincegatecontrolisusedtoinduceroutesthroughtheCNTnetwork,nano-addressesaredirectlymappedtocombinationsofgatestobeturnedonthatinduceapathfromasourcetoadestina-tion.Isthe“combinationofgates”inanywayanalogoustothenetwork,subnet,hosttypeofpartitionsthatspecifya32-bitIPv4addressofthetype151.193.204.72orthe128-bitIPv6format21DA:00D3:0000:2F3B:02AA:00FF:FE28:9C5A(or equivalent 21DA: D3: 0: 2F3B: 2AA: FF: FE28: 9C5A with leading zerosuppression)?

Doesnano-addressingrequireanentirelynewschemeforuniqueidentification?Isuniqueidentificationnecessaryatthelevelofindividualnano-addresses?Ifnec-essary,isarelativisticidentificationofinformation[160]necessary?Are“source”and“destination”comparabletoclient–serverarchitecture?

Theseandseveralotheropenquestionsarelikelytoemerge.Figure8.9illus-trates a conceptual network view of the CNT infrastructure. Superimposed onNanoComaresomeofthemedicalbenefitsthatmakeCNT-basedsensorsanddatacommunicationapowerfulallyforhealthcareimprovements.

Diabetes is not the onlydisease thatmerits prevention andmonitoring.Forearlydetectionandmonitoring,severalotherdiseasesqualifyprominently.Sensor

Molecularuser

Molecularuser

Molecularuser

Molecularuser

SensorSensor

Sensor

Glucose

NanoCom

Acetycholine

Cholesterol

Net infrastructure

DiabeticsHeart disease Gate

gate

gate

gateAlzheimer’s

Figure 8.9 NanoCom transmits data for multiple biomarkers from in vivo sen-sor nano-array. (From Bush, S.Y. and Li, Y., Nano-Communications: A New Field? An Exploration into a Carbon Nanotube Communication Network, Technical Information Series, GRC066, GE Global Research Center, Niskayuna, NY, 2006.With permission. © 2006 Interscience.)

K10373_C008.indd 286 10/8/2010 6:14:54 PM


nano-arraysmaybeoneanswerforamultifunctionaldetector,asconceptualizedinFigure8.10.Sensorsconstructedfromnanotubeschangetheirresistancebasedon the amount and specificity of the material or biomarker detected (sensed).Thus,theactofsensingmaychangetheroutingthroughtheNanoComnetwork.In other words,inastaggeredapproach,thedutycycleand/orthesleeptime[161]of thedifferentCNTnano-sensors canbe regulated to allowanyone sensor tofunction in a specified interval and detect or sense its specific analyte (glucose,cholesterol,oracetylcholine,Figure8.9).ThechangeinthepropertiesoftheCNT,whentheactofsensing is inprogress,openstheappropriate“gate,”andwhenagateisturnedon,thenanotubeswithinthegateareabecomeconducting.Then,thedataacquiredforthatspecificsensoraretransmittedtoanexternalnode.Forcomplexdiseases,wheremultiplepiecesofdataandvitalsignsmaybenecessarytomakeaninformeddecision,eitherbyhumansorinitiallybyanAI-basedintel-ligentanalyticalsystem(AGRMIA,Figure8.2),properlychoosingthesequenceofsensor,hencegates,toturnon,changesthecurrentflowtotheedgesofNanoCom,theCNTnetwork.ThelattereffectivelycreatesacontrolledNanoCom,whichmayactasorprovideweights inaneuralnetwork forAGRMIAtype systemrequir-ing a collectionofdifferentbiomarkerdata, todetermine the relative impactorrelationshipsorvaluesofvariablesthatmaybeco-integrated[162],which,iftaken

Glucose nanosensor radio

B potentostat

A

G

–60 nm

Eg

Gold nanoelectrode

Glucose, O2H2O2 Glucose oxidase

802.11b WiFi

802.11g802.16a

Grid

VISIT CLINIC

VISIT CLINIC

VISIT CLINIC

Patient data AGRMIA

Omics data Genetic risk

Conducting polymer

ECE (Pt)

(g)

Figure 8.10 Improving healthcare from the home to the hospital: “sense, then respond.”

K10373_C008.indd 287 10/8/2010 6:14:56 PM


together,maybetterreflectthestateofthepatientandthestatusofthedisease.TheMetabolicSyndromecausedbythemutationintheLRP6genemaybeanexampleofamulti-factorialdiseasewheremultipleconditionsareaffectedandmayrequiresimultaneousmonitoring.

Althoughthischapter,thusfar,mayhavesynthesizedanumberofpromisingpractical ideas, farbetterand innovative ideasandconceptsaboutproductsandservices,thantheonesmentionedhere,haveperished.Rarelyacknowledgedistheobservationthatnomatterhowgoodanideamaybe,itmayonlyshineinobscurityunless there isa strategicplanthatchartsanappropriateuseandadoptionpathincontextofothercomplementarytechnologiesandsocialawarenessofitsvalue.A commonexampleistheintroductionofthefirsthand-held“Newton,”theper-sonal digital assistant (PDA), from Apple that few may recall. Apple only sold140,000NewtonPDAsatitspeakin1993–1994andsoondiscontinuedproduc-tion.Afewyearslater,3ComintroducedthePalmPilotPDAwithfeaturesthatwere even primitive to Apple’s Newton but with the option of Internet access.Today,itmaybehardfindanyoneintheindustrialworldandprofessionalsinthedevelopingworldwhodonothaveaPDA,ofsomeformortheother.The“socialawareness”ofthevalueofPDAacceleratedwiththepenetrationoftheInternet.Newton,RIP,wasslightlyaheadofitstime.

Calls tocontain theunbridledcostofhealthcarearedemandingexplorationoftoolsandtechnologies.Theneedforwirelessremotemonitoringandtheuseofnano-biosensorsinhealthcareareasvitalasdesalinationprojects,carbonsequester-ing,metabolicengineering,cleanwater,andcleanair.Hence,innovativetoolsandtechnologiesmustalsooutlineastrategicpathfortheirintegrationandpotentialforadoptionbyhealthcaresystemswithoutexpectingacompleteoverhauloftheexistingsystem,anytimesoon.

The overall strategy for the use of wireless remote monitoring tools and themanner inwhich itmay function in the landscapeofhealthcare cost reduction(Figure8.2)isillustratedinFigure8.10.ThetimesavedbyattendingtothetwoindividualsidentifiedinFigure8.10insteadofoutpatientvisitsbysevenindividuals(Figure8.10)savescostofserviceandsuppliesandenableshealthcareprofession-alstodevotenecessarytimetothosewhoneedtheattention,henceimprovingthequality(QoS)ofhealthcare.

The normal clinical concentration of glucose in blood is in the millimolarrange,andthatprecludestheneedfornano-leveldetection.However,earlydetec-tionofotherdiseasesmayindeedbefarmoreeffectiveifdetectionispossibleatthenano-levelofproteins,peptides,molecules,ordegradedmacromoleculesthatmayholdcluestoproblemsintheirembryonicstages.Thetaskofmedicalresearchistoidentifythesemarkers,andthetaskofthetechnicalexpertsistofindwaysto identify thesemarkers through remote sensing tools.Differential profilingofgeneandproteinexpressionbetweennormalanddiseasestatesmaybehelpfultoidentifybiomarkersthatareonlyexpressedindiseasestates.Theamountsofsuch

K10373_C008.indd 288 10/8/2010 6:14:56 PM


disease-dependentbiomarkersmaybenefitfrompicoornano-leveldetectionandoffer“true”earlydetection.

Dataanddetectionascomponentsofthesystemsapproachtofuturehealth-carerequiremedicalscienceandengineeringtechnologytocreatetoolslinkedtosystems that maybe purchased, as ubiquitous generic plug-n-play commodities,fromthe localpharmacyorconvenience store inagas stationorcornergrocerystore.ConsidertherevolutionarydiscoveryoftransistorsthatproducedFETs.Thefieldhasbeenshapedbyevolutionarymarketforcesovertheyears,andFETsarenowusedaslow-costcommoditiesinthedesignofenvironmentalandagriculturalmonitoring.ItisthistrendthatFigure8.7illustrates,inconcept.ThevaluetheydeliverthroughspecificityandsensitivityisillustratedinFigure8.8.ThestrategyforintegrationandadoptionpathwayisillustratedinFigure8.10,whichincludesoneofthekey“behindthescene”driversthatmaymaterializeasmodularanalyti-calenginesoftheAGRMIAtypecollectionofresources,illustratedinFigure8.2,connectedonglobalgrids.Takentogether,remotemonitoringbyinternalwirelessnano-sensors or external sensors (as fashion rings, bracelets, wrist watches) willcoevolveanddiffuseasalifestyleapproachinhealthcare,notbecauseindividualsaresickbutbecausetheyprefertostayhealthy.Thehealthcareindustrymayben-efitfrombusinessadviceinordertotakeadvantageoftimecompression.Inotherwords,thetimetomarketfromidea(Figure8.7)toadoption(Figure8.10)maybeshorterthanthedecadesbetweenthediscoveryoftransistorsandtheuseofFETasadirtcheapcommodityforagriculturalandenvironmentalmonitoring.

Thisisthe“writingonthewall”forwhatisinstoreforremotemonitoringbywirelessnano-sensors,thenano-butlers,thatcandelivervalueatareducedcost,formicro-payments, in local andglobalhealthcare. Ignoring the suggestions inthis andother reviews or the failure to accelerate thenecessary convergence tocreatethesuggestedhealthcareservicesaswellassupportresearch[163]mayleadtochaos.

Healthcare imbalances will continue to proliferate in the United States,and severe consequences are alsopredicted forEUnations,particularly Ireland.Projectedrise inage-relatedgovernmentspendingasashareofGDPofIreland,overtheext40years,isamongthehighestintheeurozone[164].IntheabsenceofreformsandadequateinvestmentsinIreland,theadvancesininnovativeconver-gencesuggestedheremaybesluggish,atbest.But,thecostofhealthcareinIreland,whichisapproaching€3000percapita,maycontinuetoincrease(seeRef.[53]).RisingpublicdebtinIrelandmayforcespendingcutswithaconcomitantdecreaseinthequalityoflifeforthosewhoneedhealthcareservices.Costcontainmentinhealthcareservicesmaygrosslyreducepreventativemeasures,selectiveorelectiveprocedures,palliativecare,andallnonemergencyservices.Individualizedmedicineandpersonalizedhealthcarewillbeamatteroffiction,andA&Etypeacute-careservicesmaybethehealthcareskeleton.Enterprise,academia,andgovernmentcanpreventthisstateofaffairs.

K10373_C008.indd 289 10/8/2010 6:14:56 PM


8.4 Innovation Space: Molecular SemanticsTheproposalofmolecularsemantics,whetherrightorwrong,doesnotimpacttheimplementationofnano-butlers.Introducingtheconceptofmolecularsemanticsasanindependentpapermayhavebeenprudent,butthepreliminaryideameritsinclusioninthischaptertoindicatetheimportanceofstructureinmedicalscience.Currentsystems,suchasEMRSandAGRMIA,maybe,ingeneral,incapableofdealingwithstructuresunlessdedicatedprogramsareused.

8.4.1 Molecular Semantics Is about Structure RecognitionClassical semantics includesdescriptive ontologies and extractingword relation-shipsthat formbulkofthethinking[165]prevalentpresentlytomovefromthesyntactictothesemanticweb.Molecularsemanticsmaynotbenecessaryforgen-eralusagebutmayaidspecialanalyticalapplicationsinthefuturetouniquelyiden-tifymolecularorchemicalstructuresorunitsofstructuresorepitopes,inamannerthatmayhavesomesimilaritywiththeconceptofdigitalsemantics(seeRef.[40]).Uniqueidentificationofstructuresmayenablediversesystemstocomparestruc-turesinacatalogueordatabasewiththosethatmaybeidentifiedinsomediseasestates,orquery,ifanidentifiedstructurehasanyknownhomologiesorclosesimi-laritiestooneormorepartsofchemicalorbiologicalmolecules.Thesignificanceofpartialstructuresandsegmentsofstructuresorepitopesinhealthcarediagnosticsmaybebetterappreciatedfromthediscussion,laterinthissection,onautoimmunediseasesandmolecularmimicry[166].

Todelivervalueinhealthcare,partofthesolutioncallsforanalyticaltoolstoextract informationfromdata.Healthcarepresents threetypesofdatathatmayneedtoworktogether,insomecases:

1.NumericaldatainEMRSandAGRMIAplatformsmaybefeddirectlytotheanalyticalengines.

2.Syntax from patient history and physician “notes” still on paper is likelytocreatesyntaxversussemanticsnightmare if transcriptionisnecessarytocreateEMRs.The syntacticwebof today alsoplagues thebusinessworld.Thesemanticwebmovementandorganicgrowthofontologicalframeworkscontributedbyexpertsfromvariousdisciplinesarecontinuingtheirvaliantefforts to enable computer systems to “understand” and extract meaningfromsyntax.Inthiscontext,IhaveproposedaparallelapproachtoexploresemanticandontologicalframeworksusinganIPv6typeuniqueidentifica-tion scheme to enumerate data, information, and decision in a relativisticapproach(seeRef.[40]).

3.Molecularpatternsmaybepartiallynovelandespeciallyrelevantforhealth-carediagnostics.

K10373_C008.indd 290 10/8/2010 6:14:56 PM


Thissectiononmolecularsemanticsissimilaryetdistinctfrom“digitalsemantics”proposedearlier (seeRef. [40]),butbothshare theconceptofunique identifica-tiontoenableglobalsystemsinteroperabilitywithoutambiguityofidentificationorerrorsduetoambiguityinontologicalframeworksthatmaybeincomplete.Bothproposals,however,arealsolikelytobeincomplete.

Thecurrentproposalonmolecularsemanticsaddressesthethirdtypeofhealth-care-relevant data structure, that is, molecular patterns or molecular structures.Organicchemistryandbiomedicalsciencesplaceagreatdealofemphasisonpat-ternsandstructures.Hence,itmaybeworthwhiletodaretoforwardthisnewideaofhowtoenablecomputersystemstorecognizepatternsandstructuresthroughtheuseofagreedunitsofmolecularstructurethatformpartsofmacromolecules.Thisproposalfreelyborrowsideasfromseminalworksofgreatscholarsbutwithrudimentaryunderstandingoftheirdepthandwithoutanyguilt.Inadditiontoafewbuildingblocksoflinguistictheory[167],Ihavealsostretchedtheoutlinesofsemantic theoryandcognition[168],perhaps toanunnaturalandunreasonableextent.

Cognitionasrelatedtostructure innatural language ismappedtotheunitsofmolecularstructureinbiologicalmacromolecules.Ihaveextrapolatedtheideasandelementsofthelinguisticsystemtofitmolecularsemanticsinawaythatpro-posestheuseofmolecularunitsofstructures(ofmacromolecules,suchasproteins)as“agreedlexicon”tocatalyzerecognitionandunderstandingofthesestructuresbetween diverse systems to aid systems interoperability. However, in my biasedview,elementsofthelinguisticsystem(Figure8.11)seemtoresonatewiththepro-posalofmolecularsemantics.But,Ishallbethefirsttorecognizethatthisconver-gence,howeverattractive,maynotberight,intheformproposed,andadmitthatitmaybeinerror,ifitis.

Theremaybeotherways,buttheconceptofmolecularsemanticsmaybeyetanother tool to explore this scenario:MSanalysis of serum fromapatientwithunidentified typeof feverhas identified ahigh concentrationof a shortpeptidewith someambiguityabout its sequence,but itappears that thepeptidemaybepartofacommonprotein,myelin.First,aGPmayrarelysendaserumsampleforMSanalysis.Second,iftheMSdataareuploadedandanAGRMIAtypesystemisavailable,isthereatooltocomparetheMSsignaturefromthispatient(MSpat)withotherMSdatainadatabase?Thedata(MSpat)entrypointinthelinguisticsystemistheconceptualstructure,theEMRSpartoftheEMRS-AGRMIAsystem,inthiscontext.Thecomputationalresourcesthatmaybeneededtoperformthecomparativesearch,oneMSdataatatimeinthedatabase,maymakethisapproachuntenable.IfthereisacatalogueofunusualMSsignatures(MScat)intheformofa(data)dictionary,then,perhaps,itmaybefeasibletoperformthissearchanddeter-mineifamatchorcloserelationshipexistsbetweenMSpatandapatterninMScat.

Inthelinguisticsystem,thedictionaryequivalentmaybetheLexicon(Figure8.11).IsitreallynecessarytobringtheLexiconintothisdiscussion?TheMSdata

K10373_C008.indd 291 10/8/2010 6:14:56 PM


andaccompanyingexplanatorynotescouldalsoexistinarelationaldatabase,anditmaysufficeforthisexploration.TheintroductionoftheLexiconinthisschememayseem,initially,onshakygrounds,butfollowingthemodelofthelinguisticsystem,itreasonsthatiftheMSdatacatalogue(MScat)doesexistinaLexiconequivalent,itmayalsohavelinkstosyntacticstructureandsemanticstructure(Figure8.11).Thewidevariationinsyntaxneedsnoextrajustification.Hence,thedifficultytomatchsyntacticstructureswithrespecttotheLexiconevenifitcontainedamatchtoMSpatmaynotbeunexpected.Butthedata(MSpat)couldpoint todescrip-tionsthatmaybemorespecificinthesemanticstructures.ItmayidentifyMSpatasbelongingtoneuralproteins(myelinisaneuralprotein).OncetheMSpatdatamatchedtodatafromtheLexicon(MScat)pointstoneuralproteinsintheseman-ticstructure,thecorrespondencerulescouldpointbacktothesyntacticstructuresandidentifyoneormoredescriptionsofneuralproteinsthatmayoffermatchingsegments toMSpat.The realfinding ishowever thematchbetweenMSpat andthepotentiallinktomyelinthatmayhappenifanextensiveMScatismatchedbyproteinsequenceandclassifiedinthesemanticstructure.Itisquitepossiblethat

Conceptual well

formedness rules

Syntactic well

formedness rules

Conceptual structures

Syntacticstructures

Phonology

Phonetic representation

Lexicon

EMRS- AGRMIA

system

Semanticstructures

Rules of linguistic inferences

Linguistic system

Semantic well

formedness rules

Correspondence rules

Motor system

Visual system

Pragmatic

Figure 8.11 Proposed molecular semantics extrapolates concepts from the lin-guistic system.

K10373_C008.indd 292 10/8/2010 6:14:57 PM


anequivalentsetofrulesmayworkinplaceoftheRulesoflinguisticinferencetorelatetheMSdatawithMSpatandpointouttheclassofneuralproteinsandtheontologicalrelationship:myelinisaneuralprotein.TheLexicon(data)andRulesoflinguisticinference(rules)maybeworkinginconcerttoindicatethesemanticstructure,andanothersetofrules,CorrespondenceRules,couldpointtosyntaxor descriptions in the syntactic structure that builds on the identified semanticstructure.Theextentoftheclassification,precision,granularity,andotherfactorsofthesemanticandsyntacticstructurescouldbedeterminedbyrules;theequiva-lentinthelinguisticsystemaretheSemanticandSyntacticWellFormednessRules(SemWFRandSynWFR,Figure8.11), respectively.Thepresentationof thedatamayfindsomefar-fetchedrelatednesstothephoneticrepresentationandphonologydomainsinEMRS-AGRMIA,whenvisualizationofthedataisimportant.

Thusfar, Ihavenotaddressedmolecular semanticsbutattemptedtoexplorehow the linguistic system seemsmayoffer aparallel indata analysis.The latterisexpectedbecause linguisticsandartificial intelligencesharecommonelementssuchascognition,semantics,andontology.Theideaofmolecularsemanticswascamouflagedinthediscussionbecausethedata(MSpat)attemptedtofindamatchtoanexistingequivalencerelationshiptoadescriptionofrelevanceofthestructure.MolecularsemanticsindeedperformeditstaskintryingtofindamatchbetweenMSpatandMScat.ThelocationoftheMScatinthelexiconmaynotbeanoptimalexplanation,andthelexiconmayhaveadatabaseequivalentwheretheincomingdatathroughEMRS,equivalenttothevisualsystemthatfeedstheconceptualstruc-turesinFigure8.11,areformattedbycertainrules(equivalenttothePragmaticsinFigure8.11)andchanneledtoMScatthatoperatesundertheAGRMIAumbrella,whichcouldbeadatabasewithMSdata,linkedtotheLexicon(dictionary).

Molecularsemantics,thedefinitionandidentificationofstructures,maycon-tributetotheabilityofthesystemtocompareMSpattoMScatprofiles.AlthoughMSanalysisdoesnotdeliveranactualstructure,thespectraldataofferapattern,andforthisdiscussion,thispatternorprofileisreferredtoasstructure.Thiscom-parisoncanbeperformedtodayonlybyspecialapplications.

Currentsemanticwebeffortsarestrivingtostimulateglobalgroupstocontrib-utedescriptiveontologiesforchemicalandbiologicalsystemsthatmaybeacces-siblebythesoftwaretoolsandstandardspromotedbythesemanticwebexperts.Thateffortmaynotincludethepotentialforconsideringontologicalframeworksforchemicalandbiologicalstructuresordatapatterns.ToolssuchasMS,echocar-diogram,andmagnetoencephalographydonotgeneratesyntaxandthusmaybeexcluded fromontological frameworks.Thisdeficiency in thecurrentpractice isaddressedbythisproposalonmolecularsemantics.

ConservationoftheaminoacidarginineinLRP6proteinproductfromfrogstohumansmayhelpmakeitlesssurprisingtounderstand,followingthelogicofevolution, why viruses and bacteria may also share homologies with sections ofhuman DNA or why viral proteins [169] and bacterial proteins [170] may havesomehomology toaminoacid sequences found innormalandcommonhuman

K10373_C008.indd 293 10/8/2010 6:14:57 PM


proteins,suchasmyosin.Thisapparentlybenignobservationoftenproducessomedisastrousmedicalconsequences.Becausesmallsegmentsofproteinsareinvolved,thesesegmentsorepitopesmayhavestructuresthatmaybedistinct.Understandingtheform(structure)iscrucialtounderstandingthefunctioninbiologicalsystems,hencetheneedformolecularstructureinhealthcareanalysis,albeitonlyinspecialcases,andthepotentialimportanceofamechanismthatenablesanalysisofstruc-ture,thatis,molecularsemantics.

Disastrousmedicalconsequencesareobservedduetothephenomenoncom-monlyreferredtoasmolecularmimicry.Thesituationisoftencausedbyaforeignproteinfromavirusorbacteriathatshareashortstretchofaminoacidsequencehomologytoanormalhumanprotein.Theimmunesystemhappenstotargettheforeignproteinbutforsomereasonspecificallyrecognizesthishomologousregionthat servesas theantigen.Thebodyelicitsan immuneresponseandattacks theprotein.Becausethisantigenicregion(epitope)isalsoapartofanormalhumanprotein,theantibodiesproducedbythebody’simmunesystemalsorecognizetheepitopethatispresentonnormal“self”proteins.Unfortunately,theimmunesys-tembeginstodestroytheselfprotein(autoimmune)withsevereconsequencesforthepatient.Becausetheshortaminoacidsequenceoftheforeignprotein,asshortasonlysixaminoacids[171],maymimicthecorrespondingsequenceofthenormalselfprotein,thephenomenonisreferredtoasmolecularmimicry.

Thisautoimmuneresponseispartlytoblameinsomecaseswhereanindividualinalmostperfecthealthsuddenlydropsdeadfromcardiacarrestorsuccumbstoa heart attack. The etiology may be linked to common Streptococcal infection(strepthroat)thatmostchildrenandadultsexperienceatsomestageortheother.SegmentsofsomeproteinsfromStreptococcussharehomologytoproteinsspecifi-callyfoundinhearttissue[172].VariousothercasesofmolecularmimicryarewellknownincludingT-cell-mediatedautoimmunity[173]andankylosingspondylitiscausedbyonlysixaminoacids(QTDRED)foundinKlebsiella pneumoniaenitro-genaseenzyme(protein)thatexactlymatchesthehumanleukocyteantigen(HLA)receptorprotein(HLA-B27)antigenicepitope[174].

Dotheseveryshortstretchesofaminoacidscreateantigenicepitopeswithcer-tainstructuresthatmayformaclassof“superantigens”responsibleforelicitingthehumanimmuneresponsethatleadstoautoimmunediseases?Usingbasicrulesofproteinstructureandconformation,theseepitopesmayrevealstructuralpatternsthatmayinfluence“function”inbiologicalsystems,suchaselicitinganimmuneresponse.Thestructureofashortsequenceofaminoacidsmaybealmostidenticaltoanotherstructureofashortsequenceofaminoacid,butthestructuralhomologymaynotindicatethatthetwoshortstretchesofaminoacidssharesequencehomol-ogy.Computersystemsofthesyntacticwebandofthesemanticwebmayhelpinthedataanalysisthatinvolvessequence(words,suchasQTDRED)homologiesordifferencesbutmaybeincapableofdealingwithstructuresthatmaybehomolo-gouswhereassequencesarenot.ThelikelyanalyticalresultofasystempresentedwithQTDREDversusQTDREGmaysuggest,erroneously,thatthestructuresare

K10373_C008.indd 294 10/8/2010 6:14:57 PM


different.However,ifthesystem,infuture,couldusethetoolsofmolecularseman-ticsandrefertotheLexicon(Figure8.11)ofstructures,itmightrevealthatthetwostretchesofsixaminoacidsanalyzedaredifferentinsequencebutproduceidenticalstructures.Becauseantigen–antibodybindingisstructuredependentaslongasthesidechainsandgroupsaresimilar,itispossiblethattwoslightlydifferentaminoacidsequencesmayresultinnearlyidenticalstructuresthatcanstillelicitthesameautoimmuneresponse.Arelevantscenariooftenoccursinorganicchemistrywheretheempiricalformulaofachemicalorcompoundissamebutthepropertiesoftheresonantstructuresmaybedifferent.Thecurrentsemanticwebmaybeimpotenttoaddressthesestructuralissues.Incombinationwiththepowerofthesemanticweb,theremaybeaneedtoaddressstructuralissuesinhealthcarediagnostics,hencethevalueoftheproposalofmolecularsemantics.

WhydidIchoosemolecularmimicryandautoimmunediseasestomakethecaseforthevalueofthestructuralapproachintheproposalofmolecularseman-tics? The partial answer is based on the fact that almost an infinite number ofallergens and antigenic epitopes can share short homologies to self proteins. Byexploitingmolecularmimicry,itcanleadtoautoimmunediseases.Healthcare,ingeneral, and clinical immunologists, inparticular,maywish tounderstand at agreaterdepththeformandfunctionrelationshipinbiologicalsystemsandautoim-munity.Itcannotbedonewithoutstructureandstructuralcomparison,forreasonsalreadyelaboratedabove.Globalizationnowprovideswideaccesstofoodfromallaroundtheworld.Thiswindowonworldcuisinehasthepotentialtospawnnewandundocumentedformsofallergiestovariousingredientsandantigensforeigntothebody.Thepotentialtocausesomeformsofautoimmunereactionmaymanifestasinflammatoryboweldisease[175]orthegenericirritablebowelsyndrome[176].

Globalization and mobility will usher new domains of healthcare and seekknowledge about many more issues to diagnose complex diseases, for example,autoimmunitycausedbyantigens in foodproducts.Earlydetectionofantigens,which can pose the threat of molecular mimicry, may be of great significance.Researchisneededtodeterminehowtoidentifycandidatesformolecularmimicryandwhattypeofassays,invivoorinvitro,candetecttheseshortsegments.Onlytimecantellwhetherthegrowingdemandfordetectionandneedforanalysisinhealthcaremaytriggeranexplorationto“productize”molecularsemantics.Insomecases,structuralanalysismaybenecessaryorevenpivotaltocomplementnumeri-caldataandsyntactic/semanticinformationforuseinEMRS-AGRMIA(Figure8.2)typesystemsthatmustbegloballyinteroperableandlocallyresponsible.

Molecular semantics or ideas that may originate from linguistics [177] mayevolvewhenmorepeople,onbothsidesoftheaisle,medicineandengineering,canbetterappreciatethevalue,howeversubtle,offormversusfunctioninbiology.Thisproposalinitscurrentformatmaybecomeirrelevant,butitmayprovidesomecluesormayevenserveasTheGoldenKey[178]tounlockcreativityandinnovativepat-ternsinthemind[179]ofyoungpeopletodriveconvergenceofsyntax,traditionalsemantics,numericaldata,andstructure,toimproveanalysisandbenefithealthcare.

K10373_C008.indd 295 10/8/2010 6:14:57 PM


8.5 Auxiliary Space8.5.1 Potential for Massive Growth of Service

Industry in HealthcareAdoptionofthisdata-drivenmodelmayincreasethediffusionofanewclassofserviceindustry.Thegrowthofhealthservicesmaycreatenewmarketsandtrig-ger new business and revenue models (pay-per-use) even in sectors not directlyinvolvedinhealthcare,butitoffersproductsthatmaybeusedbyhealthservices,forexample,(1)softwarevendorsdeployingcloudandgridcomputingplatforms,(2) telecommunications companies billing for data transmitted in real-time,(3) dataroutingorIPconnectivityarchitectstoensureprivacy,dataconfidential-ity, and address verification using Internet Protocol version 6 (IPv6) and othersecuritytools,and(4)dataminingoutfitsthatmaycreateintelligentdifferentialdecision engines (IDDE) runningongrids, cloud computing environments, in-networkprocessing,embeddedbrowserapplications,orahostofotherplatformsyettobedeterminedordiscovered.

Thepivotalroleofdatamininginhealthcaredataanalyticsisexpectedtoevolveinwaysthatareyettobedefined.Dataminingasappliedtoso-called“businessintelligence”applicationsmayplayarolebutmaybeinadequatetoaddresstheser-vicepartofhealthcarebecausethe“service”ofhealthcareisaboutanindividualorpatient-centricdata.Ontheotherhand,thehealthcareindustrymayhavedistinc-tiveneeds,but,ingeneral,itisaboutbusinessandoperationalefficiencies.Data,information,andknowledgeholdthepotentialtoimprovebothhealthcareserviceandthehealthcareindustry,butthetoolsandapplicationsareexpectedtodifferintheirpursuitofdifferentgoalsandfunctions.Thisiswheretheprevalentviewofdataminingisexpectedtodiverge.

Thetoolsofdataminingwereenrichedwithaseaofchangeswhenprinciplesassociatedwithcomplexitytheoryandswarmintelligence[180]emergedtoofferpracticalbusinesssolutions[181]forawidevarietyofroutingandschedulingneeds.Asimilarwave(seeRef.[10])isimminentunderthegenericbannerofdatamin-ingtoolsthatmaystemfromrealitymining(seeRef.[77])anditslinkwithsocialnetworkingrelationships(seeRef.[76]).Principlesextractedfromrealityminingandsocialnetworkingparadigmsmayyieldtoolsapplicabletoavarietyoffieldsincludingbusinessservicesandhealthcareanalytics.

Datamining inhealthcare, service,and industryperspectives taken togetheralso offers a fertile ground for exploring whether the convergence of economicprinciples,tools,andtechniquesinhealthcaredataanalyticsmayleadtofurtherinnovation.Thehealthcareecosystemoffers theopportunity to test at least fourdifferenteconomicprinciplesasanalyticaltools.Forhealthcareservice,thefocusisonpatientdata,andthesedataaregeneratedbythephysiologicalsystem.Sincehumanphysiologyishighlyintegrated,itmayfollow,naturally,thatthephysiologi-calvariables(e.g.,bloodpressure,heartrate,pulserate)arelikelytobeco-integrated

K10373_C008.indd 296 10/8/2010 6:14:58 PM


(seeRef.[163]).Inotherwords,becausephysiologicalsystemsstrivetomaintainhomeostasis,itfollowsthatthegoalofphysiologyistoattainequilibrium.Whenonevariableisaffected,forexample,pulserate,itseffectis“integrated”orreflectedor related to another linked variable, for example, blood pressure. Physiologicalbalancingmechanismswithinthehumanbodywillattempttorectifythissituationandmaytrytorestorethebloodpressureoftheindividualto120/80mmHg,thenormalreading.Duetotheinnatephysiologicaldrivetorestoreequilibrium,dataanalysisinhealthcareservicemaybenefitfromapotentialexplorationandapplica-tionoftheprinciplesofNashEquilibrium[182]topredictfromasetofpatientdatawhatotherparameters(co-integrated)arelikelytochangeormaybeinfluencedbythechangedocumented(dataathand).Itmayprovidecluestoimprovediagnosis.

Dataminingtoolsforthehealthcareindustrymaybenefitfromsometraditionalapproachescoupledwithafewemergingconcepts.Likemostbusinesses,healthcareindustrysuffersfromsystemicgapsofdataandinformationinitscomplexsupplychain.Consequently,thehealthcareindustryispronetoinformationasymmetry[183]andexpectedtobenefitifinformationasymmetrycouldbereducedthroughappropriateacquisitionofdataincludingreal-timedata,forexample,fromRFID.Availabilityofhighvolumedatafromdeploymentofautomaticidentificationtech-nologies (seeRef. [131])mayhelp improve forecasting tobettermanagehumanresourcesandinventoryplanninginthehealthcareindustry.Highvolumedatamaybeinstrumentalinimprovingtheaccuracyofforecastingusingtime-seriesdataincombinationwithahostofforecastingtoolsincludingtheeconometrictechniqueofgeneralizedautoregressiveconditionalheteroskedasticity(seeRef.[46]).

Contrary to public opinion in business consulting, except in specificallydesignedbusinesscollaborations,theapplicationofNashequilibriuminbusiness(seeRef.[65])maybeconceptuallyflawed.Itislessusefulforbusinessdecisionsbutbettersuitedtohealthcareanalyticsforhealthcareservice.Incontrast,theconceptofinformationasymmetryisforeigntohumanphysiologybutisalmostasecondnaturetothecompetitivedynamicsofbusiness,whichmakesitusefulasanana-lyticaltoolinthehealthcareindustry.

TheglobalgrowthofthehealthservicesindustrythroughthemodelillustratedinFigure8.5willdwarfthecurrentrevenuesof$748billion[184]frombusinessservices.Thethreemajorglobalgiantsthatcurrentlyofferbusinessservices(IBM,HP,Microsoft),takentogether,commandlessthan10%ofthe$748billioninrev-enues.Therefore,byextrapolation,itseemsreasonabletosuggestthatsmallcom-panies,start-ups,andsmallandmediumenterprises(SME)shallfindthebarriertoentryinthehealthcareserviceindustrytobelowornil.Thehealthcareserviceindustrywill bedrivenby innovation,which isbest executedby small “skunk”worksoftalentedindividuals.Duetomultipleconvergencesnecessarytoproduceacompleteproductand/orhealthservice,corecompetencieswillbeadrivingfac-tor.Thelattermaystimulatetheneedforcollaborationandpartnershipsbetweenanumberofsmallormediumenterpriseswith localandglobalresearchinstitu-tionsandmedicalfacilities.EachgrouporallianceorSMEmaycontributeitsown

K10373_C008.indd 297 10/8/2010 6:14:58 PM


specificmoduleorcomponentbutmayfinditessentialtocooperatewithmulti-talentedteammadeupofscientists,othercompanies,medicalpersonnel,patientadvocates,andstrategiststoactasaninterfacetocatalyzeimplementation.

Stringentrequirementforahigherlevelworld-classadvisingandsupervisiontoguaranteecredibilityoftheprocess,products,andhealthservicesisnecessaryandessential.Thenoncommercial adhoc supervisory teammaybegin their involve-mentfromtheconceptualizationstageandcontinuethroughthecycleofplanning,research, product development, service creation, testing, and implementation.Criticalevaluationbyateamofacademicandcommercialexperts[185]mayhelpdefineperformanceindicators(KPI)anddeterminethestrengthofthisemergingvisionofhealthcarebyexploring(1)qualityofcareimprovements,(2)impactonhumanresourcesintermsoftimesavingsformedicalprofessional,(3)reductionincostandpotentialforsavings,(4)lengthoftimerequiredforreturnoninvestment,(5)profitabilityofbusinesses(SME)andgrowthofhighpotentialstart-ups,(6) eco-nomicbenefitsforthenation’shealthcaresystem,(7)reproducibility,portability,andsustainabilityoftheservicesmodelasaglobaltemplate,(8)businessoppor-tunitiestoimplementsimilarservicesinothercommunitiesornations,(9) creat-ingmarketalliancesinemergingeconomiestoimplementhealthcareservices,and(10)liaisonwithglobalorganizations(WorldHealthOrganization,UnitedNationsDevelopmentFund,WorldBank,AsianDevelopmentBank,Bill&MelindaGatesFoundation)tohelpintheglobaldiffusionandadoptionofhealthservicesindustrymodel.

8.5.2 Back to Basics Approach Is Key to Stimulate Convergence

The discussion in this chapter, in general, has continuously oscillated betweenmedicine, engineering, and information technology in an attempt to empha-sizeconvergenceandsuggest fruitfulanalogiesbetweenthefields.Thischapteris about data, analytics, and tools from research that may improve healthcare.Hence,theprecedingsectionsseektoharvestadvancesinsystemsengineeringandinformationcommunicationtechnologies(ICT)aswellastranslationalmedicineto improve healthcare through multidisciplinary confluence. Therefore, I shallbe ethically remiss if Idonotdigress and fail tohighlight in this sectionwhytheneedforconvergenceisacceptedbutinrealityorganizationsaresluggishtoaddress the challenges in the clinical enterprise [186]. Theproblemhas deeperimplications,andunlessreformed,theramificationsareboundtobeincreasinglydisappointing.

Initssimplestform,implementingconvergenceisofteninhibitedbythegen-eralbiomedicalilliteracyoftechnicalexpertsandtechnicalilliteracyofbiomedi-calexperts.Insightfuldegreeprogramsinbiologicalengineering[187]andhealthscience technology [188] are key mechanisms to create the supply chain of tal-ented individualswhohaveunderstandingofonefieldanddepth inanother, to

K10373_C008.indd 298 10/8/2010 6:14:58 PM


actasaknowledgebridge,whichiskeyfortheprogressofconvergence.TheU.S.physician-scientistprograms[189]thatproducegraduateswithaPhDandMDareequallyvaluableandothercountriesarebeginningtoimplementrelatedstrategies[190].However,theseprogramsonlyattractthecrèmedelacrèmeofthenation,andinsomecountriesthetotalnumberofthesehighlyqualifiedindividualsfailstoreachacriticalmass.Consequently,thefewwhosucceedoftenmovetootherpartsoftheworldwhereacriticalmassoftalentexistsandwheretheirmultipleskillsarevalued,dulyrewarded,andchallengedtoguidethenationorglobalgroups.

What is sorely needed and missing in most countries is the focus on train-ing programs for “middle level” workforce executing the bulk of the work yetremainfirmlysequesteredinonejobordomainwithoutthescopeorthedesiretobecomemultifunctional.Programswithfinancialincentives,paidleaveofabsence,structured academic training, andpractical internships arenecessary to providetechnical education for medical experts [191] and other healthcare professionals(consultants,GPs,nurses,physiotherapists,home-helpers,mentalhealthworkers)tounderstand(notnecessarilygainexpertise)thefundamentalsofmedicaldeviceengineering,sensors,remotemonitoring,communicationtechnologies,transmis-sionprotocolssuchasTCP/IP,softwarearchitecture,statistics,principlesofartifi-cialintelligence,basicprinciplesoflogic,andprogramming.Similarly,expertsinengineeringandtechnologyshouldbeofferedtheattractiveopportunity togainunderstandingofhumanphysiology,pathology,pharmacology,anatomy,cellularandmolecularbiology,neurologyandmentalhealth,genetics,principlesofinternalmedicine,nuclearmedicine,medicalimaging,biomedicaldata,inpatientandout-patientmanagementinhospitals,hierarchyofdecision-making,nutrition,social,andenvironmentalfactorsinhealth,laboratorydatareporting,andepidemiology.Cross-pollinationofideasisakeytoinnovation.

Implementingtheseparalleltrainingprogramsmaynotposeaninsurmount-ablebarrierinmostcountrieseveniftheirvisionofthefutureandcommitmenttofinanciallyinvestinitspeopleismodest,atbest.Whatislikelytosurfaceisthedifficultyofattractingsustainablenumberofcohortstotheprograms.Theproblemtoattractmaturemid-levelworkingclassforretrainingorlifelonglearning,atthetertiary level, is partially rooted in theprimary and secondary educationof thenation.Theemphasisorlackthereofonmathematicsandscienceeducationeitherdueto(1)archaicpolicies,(2)compromisedrigortofeigninclusion,(3)misguidedteachereducationprogramsthatchoosesprocessanddilutescontenttoservethelowestcommondenominator,(4)emphasisontestpreparatoryteachingwithoutroomforproblem-basedlearning,(5)inabilitytostimulateincreasingnumberoffemalestudentstotakeupadvancedmathematicsandscienceorcatalyzeyoungwomentopursuecareerpathsinthehardsciences,or(6)shoddyandsecondgradeteacher qualifications (especially in mathematics and science) masquerading asgoodenough[192].

IntheUnitedStates,aseminalreport[193]revealedthat51%ofmathemat-ics teachers intheU.S.publicK-12(primaryandsecondary)schoolsnever took

K10373_C008.indd 299 10/8/2010 6:14:58 PM


mathematicsasapartoftheircollegecurriculum.Athirdofthe“education-school-certified”scienceteachersnevertookscienceasamajorincollege.Anationalsur-vey[194]ofhighschoolphysicsfoundthat25%ofstudentstook“some”physicsinhighschooland1.2%ofseniorstudents(33,000outof2.8million)enrolledinadvancedphysics.About18%ofcertifiedteachersteachinghigh(secondary)schoolphysicshaddegreesinphysicswhile11%certifiedteachershad“degreeinphysicseducationbutnotphysics,”and27%certifiedteachersteachingphysicshadneitheradegreenoranyrelevantexperienceinthesubject.

In the Third International Mathematics and Science Study (TIMSS), theUnitedStatesranked28thinmathematicsand17thinscience,lowerthancoun-tries like Slovakia, Slovenia, Bulgaria, not to mention nations in Asia [195]. In1998,U.S.highschoolstudentsoutperformedonlytwo(Cyprus,S.Africa)oftheparticipatingcountries.Inaddition,theTIMSSclassroomstudyrevealed90%ofU.S.middleschoolmathematicslessonsareoflowqualitycomparedtoJapan(10%low)andGermany(30%low).ThequalityofmathematicsteachingwasreflectedinthepoorperformanceofU.S.eighthgraders(middleschool)reportedinthe1999TIMSSanalysis.

Thedecliningeffectivenessofmathematicsandscienceeducation is reflectedinthefactthatU.S.collegesanduniversitiesawarded24,405bachelordegreesincomputersciencein1996,50%lessthan1986(30,963in1989),andengineeringgraduatesdroppedfrom66,947in1989to63,066in1996[196].

Despitethesedisappointingtrends,theUnitedStatesisstillregardedasthe“cra-dleofinnovation”byglobalexpertsandorganizations[197].Theenigmaclearsifoneconsiderstheactualnumberofqualifiedgraduates:inthousands.Itgeneratesacriticalmassoftalenttoinnovateandcontributetoeconomicgrowth.Eachqualifiedindividualcontributesseveralmagnitudesmorethantheaveragepercapitacontri-butiontotheU.S.grossdomesticproduct(GDP).Asanexample,by1997,gradu-atesofoneU.S. institution,alone,hadfounded4000companiesemployingover1.1millionpeoplewithannualsalescloseto$250billion[198].Arecentanalysisofthesameinstitutionindicatesthatinnovationandinventionsofthisoneinstitution,annually,createnewcompaniesthatadd150,000jobsand$20billioninrevenuetotheU.S.economy,eachyear[199].Byextrapolation,thisinstitutionalone,therefore,thusfar,hascreatedcompaniesthatmaydirectlyemployover2.5millionpeopleandgenerateabouthalfatrilliondollarsinannualrevenue.Thededicationtoresearch-basedentrepreneurialspiritcoupledwiththefreedomofsomeU.S.institutionstothinkoutoftheboxaswellasthestrengthoftheU.S.investorstoassumesubstan-tialrisksarefactorsthatcontinuetoigniteinnovationandprofiteventhoughinvest-ments,bothacademicandfinancial,arenotimmunefromfailure.

Attemptsbyotherindustrializednations,withfarsmallerpopulation,topar-tiallymimictheU.S.strategyhaveproducedmixedresults.ThestrikingvisibilityoftheglobalsuccessofthegraduatesandfacultyfromU.S.researchinstitutionsincreatinginnovativecompanies,products,andservicesisbuoyedbyinvestorswill-ingtoassumegreatrisks.Inadditiontothefavorablefinancialenvironment,the

K10373_C008.indd 300 10/8/2010 6:14:58 PM


numbersorcriticalmassnecessaryforinnovationisamajordeterminanttospawnsuccess,anditmaynotbeavailableforcountrieswithlimitedpopulation.Equally,apublicbasiceducationsystemthatlacksemphasisonrigorousmathematicsandscienceeducationattheprimaryandsecondarylevelreducesthesupplychainoftalentforthefutureMDorPhDpool.Itmaybeonereasonwhymaturemid-levelprofessionalsinonefieldprefertocasta“blindeye”toconvergenceandstayintheircomfortzoneratherthanacknowledgeandtakemeasurestoimprovetheirbasicskillsinmathematicsand/orscience.Thelatterpreventsthemfromexploringtrain-ingoptionstoacquirenewdimensionsorpursuelifelonglearning,asisnecessarytocreatethetypeofmultidisciplinaryconvergencesimportantforhealthcareandhelpbuildaknowledgeeconomy.

8.6 Temporary Conclusion: Abundance of Data Yet Starved for Knowledge?

Patients want answers, not numbers. Evidence-based medicine must have num-berstogenerateanswers.Therefore,analysisofnumberstoprovideanswersistheHolyGrailofhealthcareprofessionalsanditsfuturesystems.Lackofactionduetoparalysisfromanalysisofriskassociatedwiththecomplexities[200]inhealthcareisnolongeracceptableinviewofspiralingcosts.Generatingdatawithoutimprovingthequalityofhealthcareserviceandextractingitsvalueforbusinessbenefits[201]willnotprovidethereturnoninvestment(ROI).Distributeddataandtheirrela-tionshipsaredispersedinmultiplenetworkofsystemsorsystemofsystems(SOS).Theroleofdataanalysisiscentral.ThecomatosestageoftheInformationAgeduetodataoverloadandinformationoverdoseispredictingitsdemiseunlessnewideas[202]emergeas its savior.The imminentdeathof the informationagemakes itimperativetobetterunderstandthesystemsage.Thesinglemostimportantsystemthat deserves our attention in the twenty-first century is the healthcare ecosys-tem. The convergence of characteristics such as enterprise, innovation, research,and entrepreneurship (EIRE), often common in organizations with foresight inparallelwiththevisiontodriveconvergenceofbiomedicalsciences,engineering,andinformationcommunicationtechnologies,mayactasthepurveyortoadvancehealthcarefortheprogressofcivilization[203].

AcknowledgmentNothinginthischapterwasinventedbyme.ItisonlyoriginalasadocumentbecauseIhavere-createdideasbyborrowingfromothers.Iamthecombinedeffortofpeoplewhohaveinspiredme.Hence,thischapterowespartofitsorigintoBernardLown,friend,cardiologist,andpeaceactivist.ThewisdomofCliveGranger,mathemati-cianandeconomist,andthespiritofGlennSeaborg,scientistandstatesman,were

K10373_C008.indd 301 10/8/2010 6:14:58 PM


omnipresent,too.ThatIexisttodayisduetomywife,RebeccaJaneAustinDatta,chemistandeducator,whoforciblytookmetothehospitalinMay2008,whenwewereinEurope.She“sensed”somethingwaswrong.ThemajorityofthecontentofthischapterwasdraftedwhileconvalescingintheAnaghWardofKerryGeneralHospitalinTralee,Ireland,followinganemergencycancersurgeryonMay8,2008,afewhoursafterarrivingatthehospital,inaremoteIrishhamlet.ThischapterisdedicatedtotheexceptionallyskilledsurgeonsKevinMurrayandHarisShaikhaswellasthenurses,aides,andstaffofthesurgicalandthepalliativecareunitofthehospitalinIreland.TheprofessionaltreatmentandcompassionoftheKerryGeneralHospitalteamduringmyrecoveryandsubsequentchemotherapyisashiningexam-pleofthe“artofhealing”andthetrueethosofglobalhealthcare.

References 1. Robinson,M.(2008).Healthasahumanright.European Journal of Dental Education

12(suppl1):9–10. 2. Olsen,L.A.,Aisner,D.,andMcGinnis,J.M.(2008).Engineering a Learning Healthcare

System: A Look at the Future.InstituteofMedicine,NationalAcademyofSciencesandNationalAcademyofEngineering,Washington,DC.

3. Emery,F.E.,ed.(1976).Systems Thinking.Penguin,London,UK. 4. Senge,P.M.(1990).The Fifth Discipline: Art and Science of the Learning Organization.

Doubleday,NewYork. 5. Townes, C.H. (1999). How the Laser Happened: Adventures of a Scientist. Oxford

UniversityPress,NewYork. 6. Kanigel,R.(1991).The Man Who Knew Infinity.CharlesScribner’sSons,NewYork. 7. Lewis,S.(1925).Martin Arrowsmith.JonathanCapeLtd,London,UK. 8. Sarma,S.andBrock,D.(2000).The Networked Physical World.WhitePaper.Auto-ID

Center,MIT,Cambridge,MA. 9. Krugman,P.(1994).Peddling Prosperity.W.W.Norton&Company,NewYork. 10. Finin,T.,Joshi,A.,Kolari,P.,Java,A.,Kale,A.,andKarandikar,A.(2008).Theinfor-

mationecologyofsocialmediaandonlinecommunities.Artificial Intelligence Magazine28:1–12,http://ebiquity.umbc.edu/_file_directory_/papers/376.pdf

11. Gates,B.,Myhrvold,N.,andRinearson,P.(1995).The Road Ahead.Penguin,NewYork.

12. Friedman,T.L.(2000).The Lexus and the Olive Tree.AnchorBooks,NewYork. 13. Kambil,A.andvanHeck,E.(2002).Making Markets.HarvardBusinessSchoolPress,

Boston,MA. 14. Rubin,R.E.andWeisberg,J.(2003).In An Uncertain World: Tough Choices from Wall

Street to Washington.RandomHouse,NewYork. 15. Fallows,J. (1994).Looking at the Sun: The Rise of the New East Asian Economic and

Political System.PantheonBooks,NewYork. 16. Harris,J.(2005).Homeimprovement:Changinggovernmentbusinessprocessesfor

goodwiththehelpoftechnology.Government Finance Review,April2005. 17. Joshi, Y.V. (2000). Information visibility and its effect on supply chain dynam-

ics. MS thesis, MIT, Cambridge, MA, http://www.autoidlabs.org/single-view/dir/article/6/101/page.html

K10373_C008.indd 302 10/8/2010 6:14:58 PM


18. Boot,M.(2005).Thestruggletotransformthemilitary.Foreign Affairs84:106. 19. Ranadive,V.(1999).The Power of Now: How Winning Companies Sense and Respond to

Change Using Real-Time Technology.McGraw-Hill,NewYork. 20. Finkenzeller,K.(1999).RFID Handbook: Radio Frequency Identification Fundamentals

and Applications.JohnWiley&Sons,Hoboken,NJ. 21. Heinrich,C.andBetts,B.(2003).Adapt or Die: Transforming Your Supply Chain into

an Adaptive Business Network.JohnWiley&Sons,Hoboken,NJ. 22. Cole,P.H. andEngels,D.W. (2005).21st Century Supply Chain Technology.White

Paper. Auto ID Labs, MIT, Cambridge, MA, http://www.autoidlabs.org/uploads/media/AUTOIDLABS-WP-SWNET-015.pdf

23. Fine, C.H. (1998). Clockspeed: Winning Industry Control in the Age of Temporary Advantage.BasicBooks,Cambridge,MA.

24. Caprio, D. (2003). RFID Myths and Urban Legends. White Paper. InformationTechnology Association of America, Arlington, VA, http://www.itaa.org/upload/RFID/RFID_Myths_Legends.pdf

25. Berwick,D. (2006).How tofix the system.TIME Magazine,April24,www.time.com/time/magazine/article/0,9171,1186717,00.html

26. Reich,R.B.(2001)The Future of Success.AlfredA.Knopf,NY. 27. Guyton,A.C. (1991).Textbook of Medical Physiology.W.B.Saunders&Company,

Philadelphia,PA. 28. Lehninger,A.E.(1993).Principles of Biochemistry.W.H.Freeman&Company,New

York. 29. Means,C.(2007).Riseinpatient-centriccaretoaffecthealthcare.Healthcare IT News,

February23,http://healthcareitnews.eu/content/view/148/40/ 30.Porter,M.E.(2006).Value-basedcompetitioninhealthcare:IssuesforSingapore.

November 28, http://www.isc.hbs.edu/pdf/20061128_Singapore_Health_Care.pdf

31. Schletzbaum,A. (2007).KingCountyhealthcare system emergency responseplan-ning, http://nvac.pnl.gov/meeting_spring07/presentations/Schletzbaum_Healthcare_Coalition.ppt

32. DeloitteConsulting.(2005).The Future of Healthcare: An Outlook from the Perspective of Hospital CEOs.DeloitteConsultingResearchDivision,Boston,MA.

33. TheOpenUniversity,www.open.ac.uk 34. TheUniversityofPhoenix,http://online.phoenix.edu/ 35. HealthInformationandQualityAuthority,www.hiqa.ie 36. Grimson, J., Eoghan, F., Stephens, G., Grimson, W., and Berry, D. (1997).

Interoperability issues in sharing electronic healthcare records—The synapsesapproach. In Proceedings of the 3rd IEEE International Conference on Engineering of Complex Computer Systems,pp.180–185,www.cs.tcd.ie/synapses/public/deliverables/pubs.pdf

37. Puppe,J.(2005).Transforminghealthcare.ExecutiveHealthcareManagement,http://www.executivehm.com/pastissue/article.asp?art=270556&issue=210

38. Lucas, S.M. (2007). Current state of affairs of information technology withinVHA medical centers. Testimony to the U.S. Senate Committee on Veterans’ Affairs, U.S. Congress, September 19, 2007, www.senate.gov/∼veterans/public/index.cfm?pageid=16&release_id=11324&sub_release_id=11362&view=all

39. Glaser,J.P.(2003).Clinicalinformationsystems:Thestrategyatpartnershealthcare,http://www.hisi.ie/html/ppt03/John%20Glaser.pdf

K10373_C008.indd 303 10/8/2010 6:14:58 PM


40. Datta,S.(2007).Unifiedtheoryofrelativisticidentificationofinformationinasys-tems age:Convergenceofunique identificationwith syntax and semantics throughinternetprotocolversion6(IPv6).MITEngineeringSystemsDivisionWorkingPaperSeries, Massachusetts Institute of Technology, http://esd.mit.edu/WPS/2007/esd-wp-2007-17.pdf

41. Cheekiralla, S. and Engels, D.W. (2006). An IPv6-based identification scheme.IEEE International Conference on Communications 1: 281–286. doi: 10.1109/ICC.2006.254741.

42. Datta, S., Lyu, J. and Chen, P.-S. (2007). Decision support and systems interoper-ability in global business management. International Journal of Electronic Business Management 5:255–265.MITEngineeringSystemsDivisionWorkingPaperSeries,MassachusettsInstituteofTechnology,http://esd.mit.edu/WPS/2007/esd-wp-2007–24.pdf,http://140.114.54.215/IJEBM/IJEBM_static/Paper-V5_N4/A01.pdf

43. UN/CEFACT,OASIS.(1999).ebXMLtermsofreference,www.ebXML.org 44. TheWorldWideWebConsortium,www.W3C.org 45. TheBiositemapsConsortium.(2008).NationalCentersforBiomedicalComputing,

www.ncbcs.org,www.ncbcs.org/biositemaps/Biositemaps_white_paper_v4.0.doc 46. Datta,S.,Granger,C.W.J.,Barari,M.,andGibbs,T.(2007).Managementofsupply

chain: an alternative modeling technique for forecasting. Journal of the Operational Research Society58:1459–1469.MITEngineeringSystemsDivisionWorkingPaperSeries,http://esd.mit.edu/WPS/esd-wp-2006–11.pdf

47. Simchi-Levi,D.,Simchi-Levi,E.,andKaminsky,P.(2000).Designing and Managing the Supply Chain: Concepts, Strategies and Cases.McGraw-Hill,NewYork.

48. Parlier,G.H.(2008).Transforming U.S. Army Supply Chains: Strategies for Management Innovation (in press), http://www.amazon.com/Transforming-Army-supply-chains-architecture/dp/B002I2V0Y8

49. Hilmola,O.-P.,Ma,H.,andDatta,S.(2008).Aportfolioapproachforpurchasingsys-tems: Impact of switching point. MIT Engineering Systems Division Working Paper,MassachusettsInstituteofTechnology,http://esd.mit.edu/WPS/2008/esd-wp-2008–07.pdf

50. Noll,C.(2003)InformationSystemsandtheNHSStrategy,http://www.ppa.org.uk/pdfs/bolton/information_systems_strategy.pdf

51. Drumm, B. (2006). HSE transformation programme 2007–2010. Report of theHealthServicesExecutive,http://www.hse.ie/eng/Staff/FactFile/FactFile_PDFs/Other_FactFile_PDFs/Transformation/Transformation%20Programme%202007-2010.pdf

52. NationalHealthExpenditureProjections2007–2017.(2008).CentersforMedicareand Medicaid Services, Office of the Actuary, National Health Statistics. U.S.Department of Health and Human Services (HHS), http://www.cms.hhs.gov/NationalHealthExpendData/Downloads/proj2007.pdf

53.Measuring Ireland’s Progress. (2006). Central Statistics Office, Ireland, http://www.cso.ie/releasespublications/documents/other_releases/2006/progress2006/measuringirelandsprogress.pdf

54. Datta,S.(2006).GlobalPublicGoods,#42http://dspace.mit.edu/handle/1721.1/42903 55. Fauci,A.S.,Braunwald,E.,Kasper,D.L.,Hauser,S.L.,Longo,D.L.,Jameson,J.L.,

and Loscalzo, J., eds. (2008). Harrison’s Principles of Internal Medicine, 17th edn.McGraw-Hill,NewYork.

56. Wong,D.(2006).Salivarydiagnosticspoweredbynanotechnologies,proteomicsandgenomics.Journal of the American Dental Association137:313–321.

57. UCLAHumanSalivaryProteomeProject,www.hspp.ucla.edu

K10373_C008.indd 304 10/8/2010 6:14:59 PM


58. Mulholland, A., Thomas, C.S., Kurchina, P., and Woods, D. (2006). MashUp Corporations: The End of Business as Usual. Chronicle of Service-Oriented Business Transformation.EvolvedTechnologistPress,NewYork.

59. Amendolia, S.R., Estrella, F., Hassan, W., Hauer,T., Manset, D., McClatchey, R.,Dmitry,R.,andSolomonides,T.(2004).MammoGrid:AServiceOrientedArchitecturebasedMedicalGridApplication,http://arxiv.org/ftp/cs/papers/0405/0405074.pdf

60. Google,www.google.com 61. Minsky,M.(1988).The Society of Mind.Simon&Schuster,NewYork. 62. Datta,S.(2002).Agents,#21http://dspace.mit.edu/handle/1721.1/41914 63. ClinicalDecisionMaking,http://www.medg.csail.mit.edu/ 64. Famili,A.,Nau,D.S.,andKim,S.H.,eds.(1992).Artificial Intelligence Applications in

Manufacturing.AAAIPress,MenloPark,CA. 65. Datta,S.,Betts,B.,Dinning,M.,Erhun,F.,Gibbs,T.,Keskinocak,P.,Li,H.etal.

(2003). Adaptive value network. In: Chang,Y.S., Makatsoris, H.C., and Richards,H.D. Evolution of Supply Chain Management: Symbiosis of Adaptive Value Networks and ICT,Chapter1.KluwerAcademicPublishers,Boston,MA,www.wkap.nl/prod/b/1-4020-7812-9?a=1and#3http://dspace.mit.edu/handle/1721.1/41908

66. TheMITReport. (2001).AutomatedB2Bcontracting aidedbynewgenerationofsoftwareagents,http://www.mit.edu/∼bgrosof/paps/mit-report-05-01-distrib.pdf

67. Bradshaw,J.M.,ed.(1997).Software Agents.AAAIPress,MenloPark,CA. 68. Padgham,L.andWinikoff,M.(2004).Developing Intelligent Agent Systems: A Practical

Guide.JohnWiley&Sons,Hoboken,NJ. 69. Weiss,G.,ed.(2000).Multiagent Systems: A Modern Approach to Distributed Artificial

Intelligence.MITPress,Cambridge,MA. 70. Haase,P.,Broekstra,J.,Eberhart,A.,andVolz,R.(2004).AcomparisonofRDFquery

languages,http://www.aifb.uni-karlsruhe.de/WBS/pha/rdf-query/rdfquery.pdf 71. Cesar,P.,daCosta,G.,Laskey,K.B.,andLaskeyK.J.(2005).PR-OWL:ABayesian

ontology language for the semantic web, http://mars.gmu.edu:8080/dspace/bit-stream/1920/454/1/URSW05_PR-OWL.pdf

72. Fensel,D.,Hendler,J.,Lieberman,H.,andWahlster,W.(2003).Spinning the Semantic Web: Bringing the World Wide Web to its Full Potential.MITPress,Cambridge,MA.

73. Muir,N.andKimbell,I.(2007).Discover SAP.SAPPress,Rockville,MD. 74. Stonebraker, M. (2007). Next-generation data transformation tool, http://web.mit.

edu/deshpandecenter/proj_stonebraker2.htm 75. Lanfear,D.E.,Marsh,S.,Cresci,S.,Shannon,W.D.,Spertus,J.A.,andMcLeod,H.L.

(2004).Genotypes associatedwithmyocardial infarction risk aremore common inAfrican Americans than in European Americans. Journal of the American College of Cardiology44:165–167.

76. Joshi,A.,Finin,T.,Java,A.,Kale,A.,andKolari,P.(2007).Web2.0mining:Analyzingsocialmedia.InProceedings of the NSF Symposium on Next Generation of Data Mining and Cyber-Enabled Discovery for Innovation,http://ebiquity.umbc.edu/_file_directory_/papers/379.pdf

77. Eagle,N.andPentland,A. (2006)Realitymining:Sensingcomplex social systems.Journal of Personal and Ubiquitous Computing10:255–268,http://reality.media.mit.edu/pdfs/realitymining.pdf

78. Zhu,W.,Wang,X.,Ma,Y.,Rao,M.,Glimm,J.,andKovach,J.S.(2003).Detectionofcancer-specificmarkersamidmassivemassspectraldata.Proceedings of the National Academy of Sciences (United States)100:14666–14761.

K10373_C008.indd 305 10/8/2010 6:14:59 PM


79. Lilien,R.H.,Farid,H.,andDonald,B.R.(2003).Probabilisticdiseaseclassificationof expression-dependent proteomic data from mass spectrometry of human serum.Journal of Computational Biology10:925–946.

80. Petricoin,E.F.,Ardekani,A.M.,Hitt,B.A.,Levine,P.J.,Fusaro,V.A.,Steinberg,S.M.,Mills,G.B.,Simone,C.,Fishman,D.A.,Kohn,E.C.,andLiotta,L.A.(2002).Useofproteomicpatternsinserumtoidentifyovariancancer.Lancet359:572–577.

81. Grifantini,K.(2008).Afasterwaytodetectheartattacks.Technology Review,May9. 82. Bourzac,K.(2008).Newbreath-baseddiagnostic.Technology Review,March14. 83. Denny,P.,Hagen,F.K.,Hardt,M.,Liao,L.,Yan,W.,Arellanno,M.,Bassilian, S.et al.

(2008).Theproteomesofhumanparotidandsubmandibular/sublingualglandsalivacollectedastheductalsecretions.Journal of Proteome Research7:1994–2006.

84. Biomarkers,http://www.biophysicalcorp.com/forms/Biophysical250Biomarkers.pdf 85. Velasquez-Garcia, L.F. and Akinwande, A.I. (2008). A PECVD CNT-based open

architecturefield ionizer forportablemass spectrometry. In21st IEEE International Conference on Micro Electro Mechanical Systems,Tucson, AZ, January 2008, http://eecsfacweb.mit.edu/facpages/akinwande.html

86. CardiacDiagnostics,http://www.tastechip.com/cardiac/cardiac_diagnostics_research.html

87. Ohshiro, K., Rosenthal, D.I., Koomen, J.M., Streckfus, C.F., Chambers, M.,Kobayashi, R., and El-Naggar, A.K. (2007). Pre-analytic saliva processing affectproteomic results andbiomarker screening of head andneck squamous carcinoma.International Journal of Oncology30:743–749.

88. Devadas, S. (2008). Authenticating and protecting digital information in portabledevices,http://web.mit.edu/deshpandecenter/proj_devadas.html

89. Wilson,J.S.(2005).Sensor Technology Book.Elsevier,Amsterdam,theNetherlands. 90. Rhee, S. and Liu, S. (2002). An ultra-low power, self-organizing wireless network

and its applications to non-invasive biomedical instrumentation. In IEEE/Sarnoff Symposium on Advances in Wired and Wireless Communications, WestTrenton, NJ,March13.

91. Levis,P.,Madden,S.,Gay,D.,Polastre,J.,Szewczyk,R.,Woo,A.,Brewer,E.,andCuller, D. (2004). The emergence of networking abstractions and techniques inTinyOS,http://berkeley.intel-research.net/dgay/pubs/04-nsdi-tinyos.pdf

92. Lander, E.S., Linton, L.M., Birren, B., Nusbaum, C., Zody, M.C., Baldwin, J.,Devon,K.etal.(2001).Initialsequencingandanalysisofthehumangenome.Nature409:860–921.

93. Guttmacher,A.E.andCollins,F.S.(2003).Welcometothegenomicera.New England Journal of Medicine349:996–998.

94. Moahi,K.H.(1999).HealthinformationnetworksfortelehealthinAfrica—Challengesandprospects.Libri49:43–50,http://www.librijournal.org/pdf/1999–1pp43–50.pdf

95. Grimson,J.(2008).ICT4D.TheIrish-Africanpartnershipforresearchcapacitybuild-ing.Dublin,Ireland,http://www.universitiesireland.ie/pubs/IAP-programme.pdf

96. Lown,B.(2002).Healthtechnology,thedevelopingworldandsatellife,http://www.SatlelLife.org/programs/procor/9702comm.html

97. Intel Communications Alliance. (2005). With Altobridge gateway, remote wirelesscommunicationsgetseasier,morecostefficient,http://developer.intel.com/netcomms/casestudies/Altobridge.pdf

98. World’sfirstcommercialinflightmobilephoneserviceenabledbyAltobridge’stechnology,www.altobridge.com/2010/06/11/global-times-altobridge-connecting-the-unconnected/

K10373_C008.indd 306 10/8/2010 6:14:59 PM


99. Bernard Lown’s defibrillator has jolted patients back to life for more than 40years. Technology Review November 2004, http://www.technologyreview.com/biomedicine/13894/

100. Murphy,J.G.andLloyd,M.A.,eds.(2006).Mayo Clinic Cardiology: Concise Textbook,3rdedn.MayoClinicScientificPress,NewYork.

101. Buyya,R.,Date,S.,Mizuno-Matsumoto,Y.,Venugopal,S.,andAbramson,D.(2003).Economicandondemandbrainactivityanalysisonglobalgrids,http://arxiv.org/ftp/cs/papers/0302/0302019.pdf

102. Schwindt,P.D.D.,Knappe,S.,Shah,V.,Hollberg,L.,Kitching,J.,Liew,L.-A.,andMoreland, J. (2004). Chip-scale atomic magnetometer. Applied Physics Letters 85:6409–6411,http://tf.nist.gov/general/pdf/2001.pdf

103. Negroponte,N.(1996).Being Digital.VintageBooks,NewYork.104. Kandel,E.R.andSchwartz,J.H.(1985).Principles of Neural Science.Elsevier,NewYork.105. United Nations Foundation. (2008). Case study 2: Connecting health clinics and

remote health workers (Uganda). Wireless technology for social change:Trends inmobileusebyNGOs,pp.16–18,http://www.unfoundation.org/files/pdf/2008/voda-fone/tech_social_change/health_case2.pdf

106. Truppe,M.(2005).TelenavigationwiththeVodafonemobileconnectcardfromA1.InInternational Congress for Oral and Maxillofacial Surgery,Vienna,Austria.

107. Hoyt, C.C., Richards-Kortum, R.R., Costello, B., Sacks, B.A., Kittrell, C., Ratliff,N.B.,Kramer.,J.R.,andFeld,M.S.(1988).Remotebiomedicalspectroscopicimagingofhumanarterywall.Lasers in Surgery and Medicine8:1–9.

108. Lin,N.(2005).Directcontrolofrobotsusingbrainneuralsignals.MSthesis,MIT,Cambridge,MA,http://touchlab.mit.edu/

109. Antani,S.,Demner-Fushman,D.,Li,J.,Srinivasan,B.V.,andThoma,G.R.(2008).Exploringuseofimagesinclinicalarticlesfordecisionsupportinevidence-basedmed-icine.InProceedings of the SPIE-IS&T Electronic Imaging,SanJose,CA,http://archive.nlm.nih.gov/pubs/ceb2008/2008004.pdf

110. Datta,S.(2004).Biomedicaldecisionsystems[bMDs]initiative,#52http://dspace.mit.edu/handle/1721.1/41911

111. TheVisibleHumanProject,http://www.nlm.nih.gov/research/visible/animations.html112. Weatherall,D.(1995).Science and the Quiet Art.W.W.Norton&Company,NewYork.113. Stiglitz, J.E. (2002).Globalization and Its Discontents.W.W.Norton&Company,

NewYork.114.Datta, S. (2006). Advances in SCM decision support systems: Semantic

interoperability between systems. MIT Engineering Systems Division WorkingPaper Series, Massachusetts Institute of Technology, http://esd.mit.edu/WPS/esd-wp-2006–10.pdf

115. Tuecke,S.,Czajkowski,K.,Foster,I.,Frey,J.,Graham,S.,Kesselman,C.,Maquire,T.,Sandholm,T.,Snelling,D.,andVanderbilt,P.(2003).Opengridservicesinfrastruc-ture (OGSI). Global grid forum globus toolkit 3, http://www.ggf.org/documents/GFD.15.pdf

116. Muhammad Yunus, http://nobelprize.org/nobel_prizes/peace/laureates/2006/press.html

117. Taverne, D. (2005). The March of Unreason: Science, Democracy and the New Fundamentalism.OxfordUniversityPress,Oxford,UK.

118. Goldstein, J.L. and Brown, M.S. (2001). The cholesterol quartet. Science 292:1310–1312.

K10373_C008.indd 307 10/8/2010 6:14:59 PM


119. Knowlton,A.A.andSrivatsava,U.(2008).Heat-shockprotein60andcardiovasculardis-ease:Aparadoxicalrole.Future Cardiology4:151–161.doi:10.2217/14796678.4.2.151.

120. Stone,P.H.(2007).C-reactiveproteintoidentifyearlyriskfordevelopmentofcalcificaorticstenosis:Rightmarker?Wrongtime?Journal of the American College of Cardiology50:1999–2001.

121. Cardiovasculardiseasediagnosticskit,www.roche.com/prod_diag_cardio.htm122. Bourzac,K.(2008).Next-generationdiagnostics.Technology Review,May13,http://

www.technologyreview.com/Biztech/20760/123. Rissin,D.M.,Gorris,H.H., andWalt,D.R. (2008).Distinct and long-lived activ-

ity statesof singleenzymemolecules. Journal of the American Chemical Society130:5349–5353,www.quanterix.com

124. Venter,J.C.,Adams,M.D.,Myers,E.W.,Li,P.W.,Mural,R.J.,Sutton,G.G.,Smith,H.Oetal.(2001).Thesequenceofthehumangenome.Science291:1304–1351.

125. HumanGenomeResearch,http://www.broad.mit.edu126. Fire,A.,Xu,S.,Montgomery,M.K.,Kostas,S.A.,Driver,S.E.,andMello,C.C.(1998).

Potent and specific genetic interferencebydouble-strandedRNA inCaenorhabditis elegans. Nature391:806–811.

127. ALNYLAM,http://www.alnylam.com/science-technology/index.asp128. Christensen, C.M. (1997). The Innovator’s Dilemma: When New Technologies Cause

Great Firms to Fail.HarvardBusinessSchoolPress,Boston,MA.129. David, P.A. and Wright, G. (2003). The Economic Future in Historical Perspective.

OxfordUniversityPress,NewYork.130. Datta,S.(2008).AutoIDparadigmshiftsfrominternetofthingstouniqueidentifica-

tionofindividualdecisionsinsystemofsystems.MITEngineeringSystemsDivisionWorkingPaper,MassachusettsInstituteofTechnology.SupplyChainEurope,May–June, 2008, Vol. 12(3), pp. 38–43, http://www.scemagazine.com/sce/magazines/archives/SCE_MayJune08.pdf

131. Samuel,J.-L.,Schaub,M.C.,Zaugg,M.,Mamas,M.,Dunn,W.B.,andSwynghedauw,B.(2008).Genomicsincardiacmetabolism.Cardiovascular Research79(2):218–227.

132. Zimmerli,L.U.,Schiffer,E.,Zürbig,P.,Good,D.M.,Kellmann,M.,Mouls,L.,Pitt,A.R.etal.(2008).Urinaryproteomicbiomarkersincoronaryarterydisease.Molecular and Cellular Proteomics7:290–298.

133. Villanueva,J.,Martorella,A.J.,Lawlor,K.,Philip,J.,Fleisher,M.,Robbins,R.J.,andTemps,P.(2006).Serumpeptidomepatternsthatdistinguishmetastaticthyroidcarci-nomafromcancer-freecontrolsareunbiasedbygenderandage.Molecular and Cellular Proteomics5:1840–1852.

134. Mani,A.,Radhakrishnan,J.,Wang,H.,Mani,A.,Mani,M.-A.,Nelson-Williams,C.,Carew,K.S.,Mane,S.,Najmabadi,H.,Wu,D.,andLifton,R.F.(2007).LRP6muta-tion ina familywithearlycoronarydiseaseandmetabolic risk factors.Science315:1278–1282.

135. Datta,S.,Soong,C.J.,Wang,D.M., andHarter,M.L. (1991).Purifiedadenovirus289RE1Aprotein stimulatespolymerase III transcription invitrobyaltering tran-scription factorTFIIIC. Journal of Virology 65: 5297–5304, http://dspace.mit.edu/handle/1721.1/42835

136. Datta,S.,Magge,S.,Madison,L.,andJameson,J.L.(1992).Thyroidhormonerecep-tormediatestranscriptionalactivationandrepressionofdifferentpromoters.Molecular Endocrinology 6: 815–825, http://www.jamesonlab.northwestern.edu/Bibliography/Jameson_1992_MolEndo_6_815.pdf

K10373_C008.indd 308 10/8/2010 6:14:59 PM


137. Rentoumis,A.,Chatterjee,V.K.K.,Madison,L.,Datta,S.,Gallagher,G.,DeGroot,L.J., and Jameson, J.L. (1990). Negative and positive transcriptional regulation bythyroid hormone receptor isoforms. Molecular Endocrinology 4: 1522–1531, http://dspace.mit.edu/handle/1721.1/42902

138. Chatterjee,V.K.K.,Nagaya,T.,Datta,S.,Madison,L.,Rentoumis,A.,andJameson,J.L.(1991).Thyroidhormoneresistancesyndrome:Inhibitionofnormalreceptorfunctionbymutantthyroidhormonereceptors.Journal of Clinical Investigation87:1977–1984,http://dspace.mit.edu/handle/1721.1/42900

139. Bhagwati,J.(2004).In Defense of Globalization.OxfordUniversityPress,NewYork.140. Datta,S.(2008).Canconvergenceofinnovationcatalyseeconomicgrowth?InDRIVE

Conference,#75,Kerry,Ireland,http://dspace.mit.edu/handle/1721.1/41909141. Hao,M.,Head,W.S.,Gunawardana,S.C.,Hasty,A.H., andPiston,D.W. (2007).

Direct effect of cholesterol on insulin secretion: A novel mechanism for pancreaticβ-celldysfunction.Diabetes56:2328–2338.

142. Datta,S.(2008).Informationidentificationindecisionsystems.InCIDS Conference,InstituteofTechnology,Tralee,#77http://dspace.mit.edu/handle/1721.1/41910

143. IrishMedicalOrganization:SubmissiontoBudget(2006),http://www.imo.ie/news/uploads/20061108IMO_Submission_Budget_2007final.pdf

144. Akcivi,H.A.(2003).Implementationandvalidationofareal-timewirelessnon-invasivephysiologicalmonitoringsysteminahigh-Genvironment.MSthesis,U.S.AirForceInstituteofTechnology,Wright-PattersonAirForceBase,OhioSchoolOfEngineeringandManagement(MITInstituteforSoldierNanotechnologies,http://web.mit.edu/isn).

145. Vo-Dinh,T., ed. (2007).Nanotechnology in Biology and Medicine: Methods, Devices, and Applications.CRCPress,BocaRaton,FL.

146. Forzani,E.S.,Zhang,H.,Nagahara,L.A.,Amlani,I.,Tsui,R.,andTao,N.(2004),Acon-ductingpolymernanojunctionsensorforglucosedetection.Nano Letters4:1785–1788.

147. Jensen,K.,Weldon,J.,Garcia,H.,andZettl,A.(2007).Nanotuberadio.Nano Letters7:3508–3511.

148. Rutherglen, C. and Burke, P. (2007). Carbon nanotube radio. Nano Letters 7:3296–3299.

149. Cheng,M.M.-C.,Cuda,G.,Bunimovich,Y.L.,Gaspari,M.,Heath,J.R.,Hill,H.D.,Mirkin, C.A., Nijdam, A.J.,Terracciano, R., Thundat,T., and Ferrari, M. (2006).Nanotechnologies for biomolecular detection and medical diagnostics. Current Opinion in Chemical Biology10:11–19.

150. Lin,Y.,Lu,F.,Tu,Y.,andRen,Z.(2004).Glucosebiosensorsbasedoncarbonnano-tubenanoelectrodeensembles.Nano Letters4:191–195.

151. Javey,A.,Guo,J.,Wang,Q.,Lundstrom,M.,andDai,H.J.(2003).Ballisticcarbonnanotubefield-effecttransistors.Nature424:654–657.

152. U.S.DepartmentofAgricultureReport:2006–2007.CropSeason,http://usda.mannlib.cornell.edu/usda/nass/CropProg//2000s/2007/CropProg-11-26-2007.pdf

153. Discoveryoftransistor,http://nobelprize.org/nobel_prizes/physics/laureates/1956/154. Park,T.H.,Mirin,N.,Lassiter, J.,Hafner, J.,Halas,N.J., andNordlander,P. (2008).

Plasmonicpropertiesofnanoholes.ACS Nano2:25–32,http://lanp.rice.edu/research.php155. Edwards,S.K.,Bernal-Mizrachi,L.,andRatner,L.(2006).AccumulationofNfκB1

andNfκB2isessentialforapoptosisinducedbyproteasomeinhibitioninalymphomamodel.The Ohio Journal of Science106:A14.Abstract.Scienceandengineeringonananoscale.In115th Annual Meeting,TheOhioAcademyOfScience,Dayton,OH,http://www.ohiosci.org/OJS106(1).pdf

K10373_C008.indd 309 10/8/2010 6:14:59 PM


156.Talley, C.E., Jackson, J.B., Oubre, C., Grady, N.K., Hollars, C.W., Lane, S.M.,Huser,T.R.,Nordlander,P.,andHalas,N.J.(2005).Surface-enhancedRamanscatter-ingfromindividualAunanoparticlesandnanoparticledimersubstrates.Nano Letters5:1569–1574.

157. Bush,S.F.andLi,Y.(2006).Nano-communications:Anewfield?Anexplorationintoacarbonnanotubecommunicationnetwork.TechnicalInformationSeries,GRC066.GEGlobalResearchCenter.

158. Cerf,V.G.andKahn,R.E. (1974).Aprotocol forpacketnetwork interconnection.IEEE Transactions in Communication TechnologyCOM-225:627–641.

159. Walrand,J.andVaraiya,P.(2000).High-Performance Communication Networks,2ndedn.TheMorganKaufmannSeriesinNetworking,SanFrancisco,CA.

160. Datta,S.(2007)Unifiedtheoryofrelativisticidentificationofinformationinasystemsage.#99http://dspace.mit.edu/handle/1721.1/41902

161. Karl,H.andWillig,A.(2006).Protocols and Architectures for Wireless Sensor Networks.Wiley,London,UK.

162. Granger,C.W.J. andSwanson,N.R. (1996).Furtherdevelopments in the studyofcointegratedvariables.Oxford B Economics and Statistics58:374–386.

163. O’Dowd,T.(2007).Theprivatesectorinhealthcare:Contributingtoservicebutnottoresearch.Irish Journal of Medical Science176:261–265.

164. Botman,D.andIakova,D.(2007).PolicychallengesofpopulationaginginIreland.International Monetary Fund (IMF) Working Paper WP/07/247, www.imf.org/external/pubs/ft/wp/2007/wp07247.pdf

165. Berners-Lee,T.andFischetti,M.(2000).Weaving the Web: The Original Design and Ultimate Destiny of the World Wide Web by its Inventor.Harper-Collins,New York.

166. Oldstone,M.(1987).Molecularmimicryandautoimmunedisease.Cell50:819–820.167. Chomsky,N.(2002).Syntactic Structures.Mouton,TheHague,theNetherlands.168. Jackendoff,R.(1990).Semantics and Cognition.MITPress,Cambridge,MA.169. Lane,D.andHoeffler,W.(1980).SV40largeTsharesanantigendeterminantwith

cellularproteinofmolecularweight68,000.Nature288:167–170.170. Krisher,K.andCunningham,M.(1985).Myosin:Alinkbetweenstreptococciand

heart.Science227:413–415.171. Fujinami,R.andOldstone,M.(1985).Aminoacidhomologyandimmuneresponses

betweentheencephalitogenicsiteofmyelinbasicproteinandvirus:Amechanismforautoimmunity.Science230:1043–1045.

172. Baird,R.,Bronze,M.,Kraus,W.,Hill,H.,Vasey,L., andDale, J. (1991).Epitopeof group A streptococcal M protein shared with antigens of articular cartilage andsynovium.Journal of Immunology146:3132–3137.

173. Wucherpfenning, K.W. and Strominger, J.L. (1995). Molecular mimicry inT cell-mediated autoimmunity: Viral peptides activate human T cell clones specific formyelinbasicprotein.Cell80:695–705.

174. Husby,G.,Tsuchiya,N., Schwimmbeck,P.,Keat,A.,Pahle, J.,Oldstone,M., andWilliams,R.(1998).HLA-B27-relatedantigensintissuesofpatientswithankylosingspondylitis.Scandinavian Journal of Rheumatology76(suppl):23–25.

175. Barnes,R.,Allan,S.,Taylor-Robinson,C.,Finn,R.,andJohnson,P.(1990).SerumantibodiesreactivewithSaccharomyces cerevisiaeininflammatoryboweldisease:IsIgAantibody amarker forCrohn’s disease? International Archives of Allergy and Applied Immunology92:9–15.

K10373_C008.indd 310 10/8/2010 6:14:59 PM


176. Zwetchkenbaum,J.andBurakoff,R.(1988).Theirritablebowelsyndromeandfoodhypersensitivity.Annals of Allergy61:47–49.

177. Java, A., Nirneburg, S., McShane, M., Finin,T., English, J., and Joshi, A. (2007).Using a natural language understanding system to generate semantic web content.International Journal on Semantic Web and Information Systems3:1–28,http://ebiquity.umbc.edu/_file_directory_/papers/373.pdf

178. Derbyshire,J. (2004).Prime Obsession: Bernhard Reimann and the Greatest Unsolved Problem in Mathematics.PenguinBooks,NewYork.

179. Jackendoff,R.(1994).Patterns in the Mind.BasicBooks,NewYork.180. Bonabeau,E.,Dorigo,M.,andTheraulaz,G.(1999).Swarm Intelligence: From Natural

to Artificial Systems.OxfordUniversityPress,NewYork.181. Icosystem,http://www.icosystem.com/about_management.htm182. Nash,J.F.(1950).EquilibriumpointsinN-Persongames.Proceedings of the National

Academy of Sciences (USA)36:48–49,http://www.pnas.org/misc/Nash.pdf183. Akerlof,G.(1970).Themarketforlemons:Qualityuncertaintyandthemarketmecha-

nism.Quarterly Journal of Economics84:488–500,http://nobelprize.org/nobel_prizes/economics/laureates/2001/

184. InternationalHeraldTribune,May14,2008,www.iht.com185. http://www.intota.com/expert-consultant.asp?bioID=765593&perID=722084186. Sung,N.S.etal.(2003)Centralchallengesfacingthenationalclinicalresearchenter-

prise.Journal of the American Medical Association289:1278–1287,http://jama.ama-assn.org/cgi/content/abstract/289/10/1278

187. MITDepartmentofBiologicalEngineering,http://web.mit.edu/be/index.htm188. Harvard-MITHealthSciencesandTechnology,http://hst.mit.edu189. NIHPhysicianScientist,http://www.training.nih.gov/careers/careercenter/mdphd.html190. MolecularMedicineIrelandClinicianScientistFellowshipProgramme,http://www.

dmmc.ie/MMI_Clinician_Scientist_Fellowship.htm191. Technical education for medical experts, http://www.wpi.edu/Pubs/Catalogs/Grad/

Current/bmecourses.html192. Gardner,D.P.,Larsen,Y.W.,Baker,W.O.,Campbell,A.,Crosby,E.A.,Foster,C.A.,

Francis,N.C. et al.Technical education formedical experts, http://datacenter.spps.org/sites/2259653e-ffb3-45ba-8fd6-04a024ecf7a4/uploads/SOTW_A_Nation_at_Risk_1983.pdf

193. CarnegieFoundationReport,1996.194. AmericanInstituteofPhysics,StatisticalResearchCenter.1986–1987and2000–2001

HighSchoolPhysicsSurveys,www.aip.org/statistics/trends/hs_teacher.htmlandwww.aip.org/statistics/

195. TrendsinInternationalMathematicsandScienceStudy,http://nces.ed.gov/timss196. Information Technology Association of America (ITAA) and U.S. Department of

Commerce Encourage Companies to Support Tech Education for Women andMinoritySmallBusinesses,May14,2008,http://www.itaa.org/newsroom/headline.cfm?ID=2648

197. WorldEconomicForumTechnologyPioneers,www.weforum.org/en/Communities/Technology%20Pioneers/index.htm

198. Industrial Liaison Program, Massachusetts Institute ofTechnology, http://ilp-www.mit.edu

199. MITInnovation,http://web.mit.edu/deshpandecenter

K10373_C008.indd 311 10/8/2010 6:14:59 PM


200. Bonabeau, E. (2007). Understanding and managing complexity risk. MIT Sloan Management Review48:62–68,http://www.bauer.uh.edu/2007augmgmsl7–23.pdf

201. Simchi-Levi,D.,Wu,S.D.,andShen,Z.-J., eds. (2004).Handbook of Quantitative Supply Chain Analysis: Modeling in the E-Business Era.KluwerAcademicPublishers,Boston,MA.

202. Wolpert,D.H.(2005).Information Theory: The Bridge Connecting Bounded Rational Game Theory and Statistical Physics.NASAAmesResearchCenter,MoffetField,CA.

203. Schweitzer,A. (1961).The Decay and the Restoration of Civilization.UnwinBooks,London,UK.

204. Nabel, E.G. (2003). Cardiovascular disease. New England Journal of Medicine 349:60–72.

K10373_C008.indd 312 10/8/2010 6:15:00 PM

future healthcare: bioinformatics, nano-sensors, and

Documents