applications of evolving - european respiratory society · market moves apace with new...

A version of this article appeared in the ERS Buyers’ Guide 2013/2014 under the title “Evolving technologies to study sleep and respiration: finding the dream”.

Applications of evolvingtechnologies in sleep medicine

SummaryNocturnal polysomnography (PSG) is the most important laboratory technique inthe management of sleep–wake disturbances and is considered the ‘‘goldstandard’’ [1]. New sensor technologies are entering the field, and rapiddevelopment in telecommunications and mobile technology has accelerated theintroduction of telemedicine as a viable and reliable option [2]. The present broadreview is an amalgam of the current knowledge with proposed new sensors andremote control. The reader should note that not all of the techniques discussedhere have strong clinical validation, and this should be considered whenpurchasing equipment.

State-of-the-art technology

Based on the information obtained by elec-troencephalography, electro-oculography andelectromyography (EMG), sleep stages canbe defined according to the criteria ofRECHTSCHAFFEN and KALES [3] and the newAmerican Academy of Sleep Medicine criteria[4]. Ventilation is often measured qualitativelyby means of thermistors but is more appro-priately measured with nasal pressure can-nulae, or by means of a pneumotachograph,connected with a full face mask, or calibratedinductance plethysmography, although cal-ibration of this is difficult [5]. Breathing effort

can also be detected by recording movements

of the chest and abdomen, surface EMG,

snoring, and changes in arterial blood pres-

sure, but most effectively by detection of

intrathoracic pressure swings. These swings

can be detected by measuring oesophageal

pressure. Movements of the chest and

abdomen can be recorded with strain gauges

(which detect changes in resistance accord-

ing to length changes), inductance plethys-

mography or Respitrace (with detection of

inductance characteristics of electrical con-

ductors), impedance or even a static charge

sensitive bed (which detects respiratory

movements). If respiratory effort is detected

Statement of InterestJ. Verbraecken hasacted as a consultantfor Takeda and hasreceived payments forlectures including ser-vice on speakersbureaus fromAstraZeneca, Ikaria,NVKVV, TherabelPharma, Estee Lauderand Philips. He hasreceived payment fordevelopment of edu-cational presentationsand for contribution toERS Buyers Guide fromthe EuropeanRespiratory Society.

HERMES syllabus linkmodules: B. 19

ERS 2013

Breathe | December 2013 | Volume 9 | No 6 443DOI: 10.1183/20734735.012213

Johan Verbraecken Dept of PulmonaryMedicine andMultidisciplinary SleepDisorders Centre, AntwerpUniversity Hospital,Antwerp, Belgium

Dept of PulmonaryMedicine andMultidisciplinary SleepDisorders Centre, AntwerpUniversity Hospital,Wilrijkstraat 10, 2650Edegem (Antwerp), Belgium

[email protected]

during an apnoea, this can be explained byocclusion of the upper airways. Oxygensaturation is measured by means of pulseoximetry (SpO2), as well as transcutaneouscarbon dioxide tension (PCO2). Soundrecording is another method to detectventilation. Frequency analysis of the sounds(predominantly snoring) can deliver moreinformation on flow limitation. Routinely,body position (position sensor on the chest)is also recorded. The combined applicationof these measurement techniques allowsthe assessment of normal and abnormalphysiological events in relation to sleepstructure [6].

Changing paradigm

Standard signal processing is restricted tovisual and semiautomatic analysis, whilemore in-depth information can be gatheredusing sophisticated algorithms. Updatesof current PSG systems are rare, while themarket moves apace with new technolo-gies, such as new sensors, Wi-Fi andremote monitoring (telephone or video).Integration of these new and innova-tive technologies could contribute tobetter patient care, while development ofalgorithms that can differentiate sleepstages from ECG, pulse wave, respiratorydynamics [7] and movement sensors cansimplify the diagnostic procedure. Theincrease in the processing and integrationcapacity of electronic devices, as well asadvances in low-power wireless communi-cations, has enabled the development ofunwired intelligent sensors for a wide set ofapplications.

Technologicaldevelopments in theassessment of sleep

Different approaches have been explored todevelop new technologies primarily for thediagnosis of obstructive sleep apnoea (OSA),based on more advanced analysis of ECG andpulse waves, smart assessment of arousa-bility, and new sensors. Some of these signalshave the potential to replace establishedsensors in the long term.

ECG analysis

Heart rate variability

A one-lead ECG is routinely recorded duringPSG to measure heart rate and rhythm, butlittle analysis is possible. The major challengeis very reliable R-wave detection with properidentification of ectopic beats and artefacts.Monitoring of heart rate, and in particularheart rate variability (HRV), by overnight ECGhas long been recognised as a potentialdiagnostic strategy in patients with suspectedsleep disordered breathing [8, 9]. No addi-tional electrodes or additional devices arerequired except dedicated software to carryout the analysis [10]. Analyses of the HRVcomponents within the very low-frequency(VLF), low-frequency (LF) and high-frequency(HF) domains provide insight into autonomicactivation. The VLF component concurs withvasomotor activity and thermoregulation,while the LF component corresponds withboth sympathetic and parasympathetictone and their modulation of the atrial sinusnode. In contrast, the HF component pri-marily reflects parasympathetic tone. Takentogether, the ratio of LF to HF power providesinsight into sympathovagal balance [11].Recently, some studies have shown that it iseven possible to distinguish slow-wave sleep,rapid eye movement (REM) sleep and wake-fulness based on heart rate series by using adetrended fluctuation analysis of the heartbeat [11–14].

ECG-derived respiration

Other modalities using overnight ECGinclude analysis of dynamic changes inECG morphology accompanying obstructiveevents, the accuracy of which may besignificantly enhanced by the application ofsophisticated diagnostic algorithms [15]. Thisenables the derivation of respiration from theECG by tracking the amplitudes of theprominent R-wave. The derived respiratorycurve is called ECG-derived respiration andcorrelates reasonably well with respiratoryeffort based on inductance plethysmography.The combination of ECG-derived respirationand sleep apnoea related HRV has thepotential to be a good and reliable detectiontool for sleep apnoea, but high-quality,empirical evidence on this subject is so farlacking [16].

Sleep medicine

Breathe | December 2013 | Volume 9 | No 6444

Altogether, this indicates the ECG signalis a powerful tool with rich content that caneven be used as a simplified screening device[17]. It can also be used to separate obstruct-ive from central apnoea [18].

Cardiopulmonary coupling analysis

The heart beat will slow down and speed upin tune with regular ventilation. Duringbreathing instability, the heart and breathingrates become less synchronised, and the waythe two signals are coupled together willchange. A Fourier-based analysis techniqueuses HRV and ECG R-wave amplitude fluc-tuations associated with respiration to gen-erate frequency maps of coupled autonomic-respiratory oscillations. The resulting sleepspectrogram is able to categorise sleep asstable (high-frequency coupling) or unstable(low-frequency coupling). Wake and REMsleep exhibit very low-frequency coupling(LFC). Elevated LFC is a subset of LFCthat is associated with fragmented sleep ofvarious aetiologies [19]. Cardiopulmonarycoupling measures can differentiate normalindividuals from those with moderate-to-severe OSA [20]. Incorporating cardiopulmon-ary coupling technology in routine sleepstudies could indirectly measure sleep quality.

Parameters derived from the pulse wave

Pulse transit time

The pulse transit time (PTT) is derived fromthe ECG and the photoplethysmographicarterial pressure wave measured by a fingerprobe, which has been shown to reflectinspiratory effort [21] and can therefore beused to assess cardiovascular risk. PTT isbased on measuring the time required for thearterial pulse wave to travel between twopoints in the arterial tree: from the momentwhen the pulse leaves the aortic valve (R-wave on ECG) to the time when it reaches thevessels in the finger as identified by pulseoximetry. Pulse wave speed depends on thevessel stiffness and stiffness is in turndetermined by blood pressure. Duringairway obstruction, increased swings in theintrathoracic pressure modulate blood pres-sure and induce parallel changes in PTT [1].Oscillations estimate swings of pleuralpressure that occur during the obstructivebreaths with high sensitivity and specificity.

Up-to-date PSG systems calculate PTT, but itplays a minor role in the decision tree.

Peripheral arterial tonometry and pulsewave analysis

The use of peripheral arterial tonometry(PAT) enables us to take a closer look at thepulse wave itself, which is recognised by asensor applied to the subject’s finger [22].The amplitude of the pulse wave is modu-lated by sympathetic tone and cardiovascularstroke volume. Arousals can easily be recog-nised by drops in pulse amplitude, whileslow-wave sleep and REM sleep can bedistinguished based on pulse rate analysis[23]. This method is well known as theWatchPAT device (Itamar, Cesarea, Israel).This approach has been validated in anumber of studies, whether used in isolation,or in combination with oximetry and/oractigraphy [23]. Furthermore, this approachhas the potential to allow estimation of futurecardiovascular risk via the identification ofendothelial dysfunction [24]. So far, PAT hasnot been integrated with PSG but ratherpositioned as an isolated screening tool.Currently, a compact screening device forthe wrist (SOMNOcheck micro; WeinmannMedical Technology, Hamburg, Germany) isavailable that has implemented pulse contouranalysis for this purpose. Combination of pulsewave analysis and respiratory flow signals mayallow the differentiation between obstructiveand central apnoea, and offer information onthe extent of sleep fragmentation. Recently, anovel pulse oximeter-based product has beendeveloped that uses a computer-aided scoringsolution (Morpheus Ox; WideMed, Herzliya,Israel). It uses photoplethysmography andmeasures oxygen saturation and pulse, andapplies an automatic sleep scoring software toproduce an accurate and reliable sleep ana-lysis, including identification of sleep–wake,apnoea–hypopnoea index (AHI) and Cheyne–Stokes respiration. A mobile telephone applica-tion is available, capable of recording andtransmitting the data collected during the test.This has opened a whole new dimension to theuse of oximetry to screen for sleep-relatedbreathing disorders (SRBDs).

Blood pressure

Elevated blood pressure or absence ofblood pressure dip presents the link to the

Sleep medicine

Breathe | December 2013 | Volume 9 | No 6 445

cardiovascular consequences of sleep disor-ders. Several studies have proven that OSA isan independent risk factor for developingdaytime hypertension with an increasedmorbidity and mortality [25]. In selectedcases, arterial blood pressure is measuredintermittently by a pneumatic cuff around theupper arm and a sphygmomanometer [26],but these are likely to be too ‘‘invasive’’ to beused routinely in the sleep centre because thepatient may arouse when the cuff inflates [27].In addition, the arm cuff readings cannotkeep track of the rapid blood pressurechanges that are observed in OSA and duringarousals. As an alternative, a continuousmeasurement of blood pressure can beobtained by finger photoplethysmography[28, 29]. This technique provides beat-by-beatpressure curves that allow assessment ofblood pressure variability, which may beincreased in SRBDs [30]. The finger photo-plethysmographic method consists of aminiature cuff that fits on the finger andprovides a continuous blood pressure signalderived from one (Finapres; Finapres MedicalSystems, Amsterdam, the Netherlands) ortwo inflated finger cuffs (Portapres; FinapresMedical Systems). A new approach is non-invasive blood pressure measurementwithout a cuff (based on PTT), which hasbeen clinically validated (SOMNOtouch;SOMNOmedics, Randersacker, Germany).

Capnography

The recording of PCO2 is performed usingcapnography based on infrared absorptionspectroscopy. The derived readings are end-tidal PCO2 values. The disadvantage of thismethod is impairment of patient sleepcomfort due to the need for a mask or asmall tube inserted in the exhaled air connectedto the carbon dioxide detection unit.Transcutaneous PCO2 measurement can beused as an alternative and uses heated electro-des on the skin; the electrodes have to beattached carefully because placement andsensor temperature are important modifiersof signal quality and tolerance [25]. Recently, anovel transcutaneous combined sensor forPCO2 and arterial oxygen saturation measure-ments (TOSCA 500; Linde Medical Sensors,Basel, Switzerland) was introduced into clinicalpractice [31]. Most studies concluded that thesedevices are both accurate and practicable invarious settings [31–36]. Another carbon dioxide

sensor with ultralow-power characteristicswas developed by NXP Semiconductors(Eindhoven, the Netherlands) that enablesroom temperature sensing for carbon dioxide.

Respiratory sound recordings andanalysis

Sound recording is generally performed toassess snoring. There is no widely acceptedstandard for recording sounds during sleepstudies [37, 38]. Calibrated and uncalibratedsemi-miniature microphones or vibration sen-sors taped to the skin at the level of the larynx,the suprasternal notch or forehead, and micro-phones placed at a certain distance from thehead have been employed [38–41]. For sub-sequent analysis, the microphone recordingsare only interpreted in terms of relative loud-ness and, therefore, the recording of this signaldoes not need to preserve sound signalcharacteristics [25]. A new approach tries toanalyse respiratory sounds in order to derivenoninvasive measures of increased respiratoryeffort [42]. In another system, respiratorysounds are recorded at the throat and signalprocessing separates cardiac and movementsounds from breathing sounds and snoring.Breathing is then quantified and snoringsounds are investigated in order to determinerespiratory events [43]. Combination with oxi-metry seems even more promising. Wheezemonitoring can also be useful to screen fornocturnal wheezing in patients with bronchialasthma and nocturnal respiratory symptoms.

Audiovisual recording

Continuous video monitoring is also per-formed and is particularly helpful in theassessment of nocturnal behavioural disor-ders. Time-synchronised audiovisual record-ing is part of PSG, using a video camera and aroom microphone in the sleep centre. Currentstandard infrared video cameras with infraredlight sources provide images of excellentquality [37]. Acoustic analysis could also beused for at-home screening. Finally, the useof bone-conduction microphones has thepotential to record snoring sounds directlyand reduce background noise [44].

Body and head position

Continuous recording of body position by anaccelerometer is important, as snoring and

Sleep medicine


upper airway obstruction during sleep areinfluenced by gravitation [45]. Simple sensorsencode the angle of the body into a continu-ous voltage. Other transducers use miniaturecontacts to convert an angle into a voltage ordigital code. Switch or contact-based sensorsare usually able to encode not only the anglebut also upright or supine position [25]. Thesedevices are calibrated prior to beginning thetest, identifying each voltage output with aparticular position. OSA may also be depend-ent on the position of the head. There isevidence that in a significant proportion ofpatients, trunk and head position duringsleep are not the same. Therefore, sleeprecording with dual position sensors placedon both the trunk and head should beconsidered [46]. This can be realised by anelectromechanical sensor in the middle of theforehead just above the eyebrows, mainlybased on the displacement of a smallmercury droplet within the sensor in differentpositions. This approach can be illustratedwith the Ares unicorder (Advanced BrainMonitoring, Carlsbad, CA, USA). From asingle site on the forehead, the wirelessunicorder records oxygen saturation andpulse rate (reflectance pulse oximetry), air-flow (nasal cannula), respiratory effort (usinga pressure transducer sensing foreheadvenous pressure, venous volume by photo-plethysmography and actigraphy), snoringlevels, and head movement and head posi-tion (accelerometers) [47]. ROFOUEI et al. [48]developed a new platform housed in a softneck-worn collar, composed of an ear oxi-meter sensor, a small microphone placedagainst the neck, and an accelerometermeasuring body movement and position ofthe head.

Specific movement recording

Midsagittal jaw movements

Another novel technique includes estimationof respiratory effort by measuring jaw move-ments (Jawac; Nomics, Angleur, Belgium)(fig. 1). This approach is based on magneticdistance determination [49]. A magneticsensor on the chin and another on theforehead allow the operator to continuouslydetermine the relative jaw movements, andindirectly derive respiration and snoring.Recently, it was validated against the goldstandard, the oesophageal pressure catheter,

in OSA and central sleep apnoea [50].Moreover, by an advanced analysis of thesignals, it is possible to estimate a sleep–wake profile [51]. Combination with oximetryand pulse wave analysis will allow a compre-hensive diagnosis of SRBD without full PSG.

Diaphragmatic muscle activity

Recording diaphragmatic activity from surfaceskin electrodes seems to have been neglectedfor a long time. Recently, a new easy-to-handleand noninvasive way of monitoring respiratorymuscle activity by means of surface electrodeswas introduced (Dipha; Inbiolab, Groningen,the Netherlands), which can be used asan indirect measure to assess respiratorymechanics during wake and sleep [52] (fig. 2).

Temperature

New flow sensor

A new type of thermal sensor using poly-vinylidene fluoride (PVDF) film (DymedixCorp., Minneapolis, MN, USA) has beendeveloped with a faster response timethan those of traditional thermal devices.Thermocouples have a response time of,1 s, whereas the PVDF response time ison the order of 0.005 s. The signal producedby PVDF film is 150 000 times stronger thanthe signal produced by a thermocouple andappears to estimate changes in airflow moreaccurately. The PVDF signal is proportional to

Figure 1Jawac technology: sensing mandibular movementsduring sleep. Reproduced with permission fromNomics (Angleur, Belgium).

Sleep medicine


the difference in temperature between thetwo sides of the film. A single PVDF devicecan detect both nasal and oral flow [53, 54].

Body temperature and metabolic activity

Temperature regulation in the body can alsobe used to monitor sleep quality. Skintemperature increases during sleep onsetand decreases during wakeup [55].

Core body temperature measured from anindwelling rectal probe provides informationon the internal circadian rhythm generator(clock) [56] and gives insights on the actualcircadian phase of the patient. Core bodytemperature is closely linked to the circadiansystem and its recording will allow theevaluation of delayed or advanced sleepphase problems. The normal differencebetween maximum and minimum bodytemperature in the diurnal rhythm is closeto 0.5uC. Rectal or ear temperature probes aremost appropriate [25]. The sleep switchdevice is a small device held between theindex finger and thumb, the subject’s gal-vanic skin response closing the electric circuitbeing scored as wake. Subjects falling andbeing asleep would release the pressure fromthe device and open the electric circuit [57].The Q Sensor (Affectiva, Waltham, MA, USA)is a wireless logging biosensor that measuresskin conductance, skin temperature andmotion comfortably from the wrist [58]. A

taxonomy of autonomic sleep patterns will bedeveloped based on electrodermal activity.Integration of movement sensors with tem-perature and (eventually) humidity sensors,will allow the differentiation of the sleepstages, based on new algorithms. TheSenseWear BMS sensor system (BodyMediaInc., Pittsburgh, PA, USA) is designed tocontinuously monitor energy expenditure,motor activity and sleep efficiency. Severalsensors are incorporated into a single devicethat is worn on the back of the upper rightarm over the triceps muscle and held in placeby a Velcro armband. The sensors are a two-axis accelerometer, a heat flux sensor, agalvanic skin response sensor, a skin tem-perature sensor and a near-body ambienttemperature sensor. Data from these para-meters are used to estimate energy expend-iture using proprietary equations developedby the manufacturer [59]. A simple system hasbeen described that continuously measures thebody temperature at the ear with an electronictemperature sensor that is coupled to aBluetooth transmitter that sends the data to amobile phone where the data are stored [60].

Smart noncontact systems

Image processing technologies and mattresssensor techniques are proposed as newrespiratory monitoring techniques which donot restrain the patient.

Intercostals

Intercostals

Common

Electrodepositions

Respiratory EMG

Diaphragm F

Diaphragm F

Diaphragm D

R6

L

LR

1

3 4

2

5

RMS Interc

RMS F Dia

Figure 2Dipha device for transcutaneous diaphragmatic muscle activity. Reproduced with permission from Inbiolab(Groningen, the Netherlands). RMS: root mean square of the myoelectric signal.

Sleep medicine


Doppler radar technique

A potentially attractive alternative approachinvolves the use of a noncontact device(SleepMinder; BiancaMed, Dublin, Ireland) setup next to the patient’s bed [8]. It uses thereflection of radio waves to derive sleep,respiratory effort, body movement and evencardiac movement from a distance via asophisticated signal analysis algorithm [61] thatseparates these signal components. Thisenables the identification respiratory events.The available data suggest this device mayprove to be a useful addition to the currentdiagnostic armamentarium [62], with the AHIcorrelating strongly with PSG measurement [63].

Thermal infrared imaging

Measurement of breathing rates is feasible ata distance using thermal infrared imaging.This method is based on the presence orabsence of a hot expiratory plume in thevicinity of the nostrils. When analysing thethermal imaging signal, the full breathingwaveform can be extracted (rate and ampli-tude). Consistent segmentation of the nostrilarea and facial-tracking algorithms enablesustained monitoring of breathing, resultingin functionality equivalent to that of athermistor, delivered in a contact-free manner[64]. The centrepiece of the system is amidwave infrared camera and the acquiredimaging signals are processed by real-timecustom software.

Passive infrared technology

The breath motion detecting system (CHRNamur, Namur, Belgium) was developedbased on passive infrared (PIR) sensortechnology for a contactless detection ofrespiratory movements. This sensor is adual-element pyroelectric sensing device thatreacts only to heat source variation (such asthe movement of the human body). Allobjects emit infrared radiation that is invis-ible to the human eye but can be measured byelectronic devices. When exposed to infraredradiation, the material of the pyroelectricsensor generates a surface electric charge.The term passive in this instance means thatthe PIR device does not emit an infraredbeam but merely passively accepts incominginfrared radiation. The ideal distance appearsto be 20–50 cm between the patient and the

sensor. The pilot study in 169 patientsdemonstrated the ability of the PIR techno-logy to detect respiratory movements inadults [65].

Fibre-grating vision sensor

A fibre-grating vision sensor is a noncontactrespiratory monitoring system to detectchanges in volumes by measuring the move-ment of laser spots on the body surface.The system measures the up-and-downmotion in o100 sampling points on theupper half of the body and outputs the sumas respiratory movement [66]. The systemis able to discriminate OSA from centralsleep apnoea.

Mattress sensor technology

The static charge-sensitive mattress (BioMatt;Biorec, Turku, Finland) was developed toanalyse body movements, respiration andcardiac pulses (ballistography) by recordingthese with a very sensitive pressure mattressthat was placed underneath the actual mat-tress [67]. These devices were not widelyadopted, except in Scandinavian countries.

Recently, a smaller (3266260.4 cm)electromechanical film transducer sensor(Emfit, Vaajakoski, Finland) was validatedthat is placed under the thoracic area of thesleeping patient. Based on advanced signalprocessing, respiratory effort, AHI and sleepstages could be calculated [68].

KOGURE et al. [69] developed a highlysensitive pressure bed sensor that is placedunder a mattress. It can identify an ‘‘in-bed/

out-of-bed’’ state from the vibrations of themattress. Combined with wrist actigraphy, itenables convenient long-term sleep-relatedevaluation.

MERILAHTI et al. [70] proposed the bedoccupancy sensor, which consists of a thinpressure-, light- and temperature-sensitivelayer used as an additional sensor to actigra-phy. The use of pressure-sensitive sensorsinstalled in a bed enables respiration monitor-ing during sleep, employing the combineduse of a pressure sensor, and a multipointelectric sensors (polypiezofilm sensing) andpneumatics-based methods that use a thin,air-sealed cushion and a mattress. In addition,silicon rubber tubes and plastic opticalfibre sensors have been proposed due totheir flexibility, light weight, immunity to

Sleep medicine


electromagnetic interference and ease ofinstallation in a bed. These sensors can detectbending on an optical intensity basis [71].

Technologies based on air tubes at thesame initial pressure, connected with a sens-itive differential pressure sensor, were vali-dated in a pilot study [72]. Both tubes areplaced in parallel between the subject’sshoulders and hips underneath a mattresstopper. The signal obtained with the sensorcan be subdivided into a respiratory and a heartrate signal by means of a Butterworth band-pass filter. Additionally, the signal is processedwith a differential and integration filter, fol-lowed by peak detection, which enables thedetection of heart rate and breathing frequencyin an unobtrusive manner [72].

ZHU et al. [73] used a similar system basedon polyvinyl tubes, 30 cm in length and4 mm in diameter, filled with air-free waterpreloaded at an internal pressure of 3 kPa.One end was connected to a liquid pressuresensor head, and enabled derivation of a

static pressure component corresponding tothe weight of the head and a dynamicpressure component reflecting changes dueto respiratory motion and heart beats [73].

Several other authors have presentedresults on detecting heart rate and breathingrate from signals off-body, using air tubesplaced below the thorax or pillow, piezo-polymer sensor mats or force sensors in thebed legs [74–76]. These sensors avoid dis-turbing the patient during sleep and allow forlong-term monitoring, even in the homesetting, at the same time.

MIGLIORINI et al. [77] performed HRVanalysis and automatic sleep stage classifica-tion through bed sensors. However, commer-cially available pressure mats are unableto accurately reproduce indentation anddeformation based on the measurement ofpressure distributions.

The IdoShape (Custom8, Leuven, Belgium)is a sensor mat that detects movement andmeasures real deformation of the mattresssurface, and can be used to evaluate the fitbetween the human body and its biomechani-cal support. The mCliMat (Custom8) sensormat monitors temperature and humidity in amattress system using 512 sensors, andgenerates thermodynamic analyses of heatflow and moisture transport (fig. 3).

BRINK et al. [76] reported the application offour high-resolution force sensors installedunder the bedposts. The recoil movement ofthe body at each heartbeat, as well as the liftingand lowering of the thorax while breathing,causes very small shifts of the centre of gravityof the bed and the subject. These shifts arereflected in the altering force distributionsacross the four sensors. The signals from thedifferent load cells are combined to create abreathing signal and, again, cardiac andrespiratory parameters can be calculated.

BEATTIE et al. [78] demonstrated the feas-ibility of using unobtrusive load cells installedunder the bed to measure AHI. Calibrated cellscan even be used to monitor a patient’s weightas well as the lying position of an individual.Similar studies with bed-installed load-cellsensors were performed to derive HRV.

Intelligent textile technology

Smart garments

Textile technology was proposed as aninnovative tool for the development of

Temperature

Absolute humidity

Figure 3mCliMat sensor mat monitors temperature and humidity. Reproduced with permissionfrom Custom8 (Leuven, Belgium).

Sleep medicine


devices for vital sign monitoring [79]. Inparticular, textile sensors can now be embed-ded in smart garments for collecting a varietyof biological signals (ECG, bioimpedance,skin resistance, respiratory frequency, etc.).Prototypes of smart garments are availablein various research laboratories around theworld. For details, see the review by LYMBERIS

and DITTMAR [80]. From a methodologicalpoint of view, the development of thesedevices implies the minimisation of theweight and size of electronics, reduction ofpower consumption, and design of specificalgorithms for signal conditioning. However,the most crucial issue to be addressedin dealing with smart garments is the min-imisation of motion artefacts. The SD-101(Kentzmedica Co. Ltd, Saitama, Japan) is anonrestrictive, sheet-like medical device withan array of pressure sensors to detect SRBDby sensing gravitational alterations in thebody corresponding to respiratory move-ments [81].

Conductive fibres

The SD-101 vest also integrates a textile-based piezoresistive plethysmograph thatdetects changes in the thorax circumferencefrom which respiratory frequency is derived.This plethysmograph was obtained by aprocessing of the conductive fibres.Pathways made of the same conductive fibresconnect the above transducers with theelectronic module, which is hooked to thevest at waist level by a Velcro strip. Theelectronic module includes a three-axis accel-erometer to detect subjects’ movements,stores data on a local memory card, andcan transmit all signals via Bluetooth to anexternal computer for data visualisation,storage on disk and, possibly, a relaytransmission through Wi-Fi or universalmobile telecommunications system connec-tions to a remote monitoring station [79, 82].

Telemedicine

Application of telemedicine in SRBD allowssupport of the patient in order to improvetreatment compliance, to remotely downloadpreviously recorded data [83–87] or to transmitreal-time physiological parameters whilethe patient is sleeping (fig. 4) [88, 89].Telemonitored PSG has the promise to over-come the disadvantages of home recordings

and could provide an organisational solutionto the overloading of specialised sleep centres.Technicians can regularly verify, at a distance,the quality of the PSG recordings by meansof periodic access to the PSG monitoringdevice. From their telemonitoring controlpanel, they are able to insert comments inthe recording, adjust transducer gain and,in the event of a technical accident, informthe patient by telephone [90]. GAGNADOUX

et al. [91] reported that PSG performed in alocal hospital and telemonitored by a sleeplaboratory was clearly superior to unattendedhome PSG. The major problem encounteredwith home sleep studies is the potentialloss of data in 4.7–20% of the cases [92],which leads to less cost savings thanexpected. Using the Sleepbox technology(Medatec, Brussels, Belgium), which is awireless system able to communicate withthe polysomnograph and with the Internetthrough a Wi-Fi/3G interface, and makinguse of communication via Skype (Microsoft,Redmond, WA, USA), the authors wereable to deliver recordings of excellentquality in 90% of the cases. PELLETIER-FLEURY

et al. [90] comparatively evaluated the costand effectiveness of PSG telemonitoring andPSG by conventional unsupervised homemonitoring, demonstrating that remote tele-monitoring made the procedure clearly super-ior from a technical point of view and waspreferred by the patients. They calculated acost of USD 244 for PSG telemonitoring,whereas the PSG with conventional unsuper-vised home monitoring cost USD 153.Healthcare infrastructure savings also have tobe taken into account. If we add up, forinstance, the travelling costs avoided by thepatients and the working days that they did notlose, we could estimate that the real cost wouldbe similar to or lower than that of conventionalPSG [93]. In any case, remote monitoringopens an interesting perspective to decreasethe failure rate of home sleep studies. KRISTO

et al. [94] presented a telemedicine protocol forthe online transfer of PSGs from a remote siteto a centralised sleep laboratory. Their systemwas based on the transmission of data usingFTP (file transfer protocol). SEO et al. [95]developed a nonintrusive home-based health-monitoring system to monitor patients’ ECGresults, weight, movement pattern and snor-ing. CHOI et al. [96] presented a ubiquitoushealth monitoring system in a bedroom, whichmonitors ECG, body movements and snoring

Sleep medicine


with non-conscious sensors. A centralisedframework is used in most home telehealthsystems, in which a centralised databaseis used for data storage and analysis [97].In 2008, KAYYALI et al. [98] proposed a newcompact telemetry-based sleep monitor(PSG@Home), consisting of a 14-channelwearable wireless monitor and a mobilephone-base gateway to transfer data, includingvideo, in real time from the patient’s home to aremote sleep centre. The receiver is a separateunit attached to the back of the display.Internal Bluetooth receivers, often included inmany laptops, can also be used instead of adedicated external Bluetooth receiver.

Mobile telephones

Mobile telephones are a very attractiveapplication platform, as most of them incorp-orate a lot of interesting characteristics,namely computing power, long battery life,many built-in sensors and wireless connectiv-ity. Mobile applications may also simplifysleep diagnostics by using a smartphone’sbuilt-in sensors. The parameters that can bemeasured by such applications are breathingpattern and movement patterns, which arerecorded using the built-in microphone andaccelerometer, respectively. Consequently,the recorded data are sent to a server foranalysis in dedicated software like MATLAB(MathWorks, Natick, MA, USA) in order to

diagnose patients [99]. Sleep Cycle (MaciekDrejak Labs, Gothenburg, Sweden) is apopular iPhone (Apple Inc., Palo Alto, CA,USA) ‘‘app’’ that uses the accelerometer inthe iPhone to monitor body movements anddetermine the user’s sleep stage. The userjust needs to put the iPhone in a suitableplace on the bed. However, the iPhone canaccidently fall off the bed and it needs to beconnected to the charger for the whole night.Tablet computers or mobile telephones alsoenable an online signal check by sending ascreenshot of the recording. Transfer timesand periods can be programmed individually.

Conclusions

Traditional sleep monitoring methods use avariety of leads and probes on the patient’sface and body to gather data. Additionalinformation can be achieved from thesesignals by advanced processing based oncomplex algorithms. Moreover, a number ofsignals that are not traditionally used inclinical PSG will become of interest forspecific patient categories. We are also facedwith the development of innovative noncon-tact systems based on movement detectionusing radar and infrared technology. The ideaof automatic sleep evaluation and monitoringthrough signals that are integrated into theenvironment (a sensorised bed) or through

PatientPersonal

dataprocessing

unit

Home computer

ServerSignal

analysis

Clinician’s computer

Figure 4Model for telemonitoring in sleep medicine. Data flow is initiated at the body area network of sensors on the leftand eventually reaches the clinician’s workstation on the right.

Sleep medicine


wearable textile technology will changethe traditional paradigm of clinical poly-somnography. Implementation of wirelessapplications and remote monitoring will lead

to new platforms and evolve towards low-threshold sleep telemedicine. The availableevidence base has, however, lagged farbehind.

References

1. Riha R. Polysomnography. In: Simonds A, De BackerW. ERS Handbook: Respiratory Sleep Medicine.Sheffield, European Respiratory Society, 2012;pp. 120–130.

2. Sagahyroon A, Aloul F, Al-Ali AR, et al. Monitoringpatients’ signs wirelessly. J Med Imag Health Inf 2011;1: 252–55.

3. Rechtschaffen A, Kales A. A manual of standardisedterminology, techniques and scoring for sleep stagesof human subjects. Los Angeles, Brain informationService/Brain Research Institute, 1968.

4. Iber C, Ancoli-Israel S, Chesson AL, et al. The AASMManual for the Scoring of Sleep and AssociatedEvents. Rules, Terminology and TechnicalSpecifications. 1st Edn. Westchester, AmericanAcademy of Sleep Medicine, 2007.

5. Sleep-related breathing disorders in adults. recom-mendations for syndrome definition and measure-ment techniques in clinical research. The report of anAmerican Academy of Sleep Medicine Task Force.Sleep 1999; 22: 667–689.

6. Bloch KE. Polysomnography: a systematic review.Technol Health Care 1997; 5: 285–305.

7. Chung GS, Choi BH, Lee JS, et al. REM sleepestimation only using respiratory dynamics. PhysiolMeas 2009; 30: 1327–1340.

8. Kent B, McNicholas WT. Ambulatory investigation ofsleep-disordered breathing. Breathe 2011; 8: 107–112.

9. Guilleminault C, Connolly S, Winkle R, et al. Cyclicalvariation of the heart rate in sleep apnoea syndrome.Mechanisms, and usefulness of 24 h electrocardio-graphy as a screening technique. Lancet 1984; 1:126–131.

10. Penzel T, McNames J, de Chazal P, et al. Systematiccomparison of different algorithms for apnoeadetection based on electrocardiogram recordings.Med Biol Eng Comput 2002; 40: 402–407.

11. Decker MJ, Eyal S, Shinar Z, et al. Validation of ECG-derived sleep architecture and ventilation in sleepapnea and chronic fatigue syndrome. Sleep Breath2010; 14: 233–239.

12. Bulckaert A, Exadaktylos V, De Bruyne G, et al. Heartrate based nighttime awakening detection. Eur J ApplPhysiol 2010; 109: 317–322.

13. Bulckaert A, Exadaktylos V, Haex B, et al. Elevatedvariance in heart rate during slow-wave sleep afterlate-night physical activity. Chronobiol Intern 2011; 28:282–284.

14. Bunde A, Havlin S, Kantelhardt JW, et al. Correlatedand uncorrelated regions in heart-rate fluctuationsduring sleep. Phys Rev Lett 2000; 85: 3736–3739.

15. Heneghan C, de Chazal P, Ryan S, et al.Electrocardiogram recording as a screening tool forsleep disordered breathing. J Clin Sleep Med 2008; 4:223–228.

16. Collop N, Tracy SL, Kapur V, et al. Obstructive sleepapnea devices for out-of-center (OOC) testing:technology evaluation. J Clin Sleep Med 2008; 7:531–548.

17. Heneghan C, Chua CP, Garvey JF, et al. A portableautomated assessment tool for sleep apnea using acombined holter-oximeter. Sleep 2008; 31: 1432–1439.

18. Gubbi J, Kandoker A, Palaniswami M. Classificationof sleep apnea types using wavelet packet analysis ofshort-term ECG signals. J Clin Monit Comput 2012;26: 1–11.

19. Schramm PJ, Thomas R, Feige B, et al. Quantitativemeasurement of sleep quality using cardiopulmonarycoupling analysis: a retrospective comparison ofindividuals with and without primary insomnia. SleepBreath 2013; 17: 713–721.

20. Harrington J, Schramm PJ, Davies CR, et al. Anelectrocardiogram-based analysis evaluating sleepquality in patients with obstructive sleep apnea. SleepBreath 2013 [In press DOI: 10.1007/s11325-013-0804-9].

21. Argod J, Pepin J, Levy P. Differentiating obstructiveand central sleep respiratory events through pulsetransit time. Am J Respir Crit Care Med 1998; 158:1778–1783.

22. Bar A, Pillar G, Dvir I, et al. Evaluation of a portabledevice based on peripheral arterial tone for unatten-ded home sleep studies. Chest 2003; 123: 695–703.

23. Dvir I, Adler Y, Freimark D, et al. Evidence for fractalcorrelation properties in variations of peripheralarterial tone during REM sleep. Am J Physiol HeartCirc Physiol 2002; 283: H434–H439.

24. Sommermeyer D, Schwaibold M, Scholler B, et al.Prediction of cardiovascular risk from peripheralpulse wave. IFMBE proceedings. World Congress onMedical Physics and Biomed Eng. Vol 25/7. Springer,Berlin, 2009; pp. 891–892.

25. Penzel T, Canisius S. Polysomnography. In:Randerath WJ, Sanner BM, Somers VK, eds. SleepApnea. Basel, Karger, 2006; 35: pp. 51–60.

26. Hoffstein V, Rubinstein I, Mateika S, et al.Determinants of blood pressure in snorers. Lancet1988; 2: 992–994.

27. Hirshkowitz M, Kryger MH. Monitoring techniquesfor evaluating suspected sleep-disordered breathing.In: Kryger MH, Roth T, Dement WC. Principles andPractice of Sleep Medicine. 4th Edn. Philadelphia,Elsevier Saunders, 2005; pp. 1378-1393.

28. Mateika JH, Mateika S, Slutsky AS, et al. The effect ofsnoring on mean arterial blood pressure during non-REM sleep. Am Rev Respir Dis 1992; 145: 141–146.

29. Davies RJO, Crosby J, Vardi-Visy K, et al. Non-invasive beat to beat arterial blood pressure duringnon-REM sleep in obstructive sleep apnea. Thorax1994; 49: 335–339.

30. Davies RJO, Vardi-Visy K, Stradling JR. Identificationof sleep disruption and sleep disordered breathingfrom the systolic blood pressure profile. Thorax 1994;48: 1242–1247.

31. Parker SM, Gibson GJ. Evaluation of a transcuta-neous carbon dioxide monitor (‘‘TOSCA’’) in adultpatients in routine respiratory practice. Respir Med2007; 101: 261–264.

Sleep medicine


32. Severinghaus JW, Stafford M, Bradley AF. tcPCO2

electrode design, calibration and temperature gra-dient problems. Acta Anaesthesiol Scand Suppl 1978;68: 118–122.

33. Bendjelid K, Schutz N, Stotz M, et al. TranscutaneousPCO2 monitoring in critically ill adults: clinicalevaluation of a new sensor. Crit Care Med 2005; 33:2203–2206.

34. Senn O, Clarenbach CF, Kaplan V, et al. Monitoringcarbon dioxide tension and arterial oxygen saturationby a single earlobe sensor in patients with criticalillness or sleep apnea. Chest 2005; 128: 1291–1296.

35. Bolliger D, Steiner LA, Kasper J, et al. The accuracy ofnon-invasive carbon dioxide monitoring: a clinicalevaluation of two transcutaneous systems.Anaesthesia 2007; 62: 394–399.

36. Stege G, van den Elshout FJJ, Heijdra YF, et al.Accuracy of transcutaneous carbon dioxide tensionmeasurements during cardiopulmonary exercisetesting. Respiration 2009; 78: 147–153.

37. Vaugh BV, Giallanza P. Technical review of poly-somnography. Chest 2008; 134: 1310–1319.

38. Hoffstein V. Snoring. Chest 1996; 109: 201–222.39. Hoffstein V, Mateika S, Anderson D. Snoring: is it in

the ear of the beholder? Sleep 1994; 17: 522–526.40. Dalmasso F, Righini G, Fava C, et al. Snoring:

measurement and analysis. Monaldi Arch Chest Dis1993; 48: 589–595.

41. Dalmasso F, Benedetto G, Pogolotti R, et al. Digitalprocessing of snoring sounds. Eur Respir J 1990; 11:528–532.

42. Kaniusas E, Pfutzner H, Saletu B. Acoustical signalproperties for cardiac/respiratory activity and apneas.IEEE Trans Biomed Eng 2005; 52: 1812–1822.

43. Yadollahi A, Moussavi Z. Acoustic obstructive sleepapnea detection. Conf Proc IEEE Eng Med Biol Soc2009; 2009: 7110–7113.

44. Nasu Y, Ashida N, Yamakawa M, et al. Automaticidentification of apnea through acoustic analysis forat-home screening. Telemed J E-Health 2011; 17:467–471.

45. Oksenberg A, Silverberg DS. The effect of bodyposture on sleep-related breathing disorders: factsand therapeutic implications. Sleep Med Rev 1998; 2:139–162.

46. van Kesteren ER, van Maanen JP, Hilgevoord AA,et al. Quantitative effects of trunk and head positionon the apnea hypopnea index in obstructive sleepapnea. Sleep 2011; 34: 1075–1081.

47. Levendowski DJ, Morgan T, Montague J, et al.Prevalence of probable obstructive sleep apnea riskand severity in a population of dental patients. SleepBreath 2008; 12: 303–309.

48. Rofouei M, Sinclair M, Bittner R, et al. A non-invasivewearable neck-cuff system for real-time sleep mon-itoring. Proc Int Conf Body Sensor Networks 2011;156–161.

49. Senny F, Destine J, Poirrier R. Midsagittal jawmovement analysis for the scoring of sleep apneaand hypopneas. IEEE Tran Biomed Eng 2008; 55:87–95.

50. Maury G, Cambron L, Jamart J, et al. Added value of amandible movement automated analysis in thescreening of obstructive sleep apnea. J Sleep Res2013; 22: 96–103.

51. Senny F, Destine J, Poirrier R. Midsaggital jawmovements as a sleep/wake marker. IEEE TransBiomed Eng 2009; 56: 303–309.

52. Duiverman ML, van Eykern LA, Vennik PW, et al.Reproducibility and responsiveness of a non-invasiveEMG technique of the respiratory muscles in COPDpatients and in healthy subjects. J Appl Physiol 2004;96: 1723–1729.

53. Berry BB, Koch GL, Trautz S, et al. Comparison ofrespiratory event detection by a polyvinylidenefluoride film airflow sensor and a pneumotachographin sleep apnea patients. Chest 2005; 128: 1331–1338.

54. Koo BB, Drummond C, Surovec S, et al. Validation ofa polyvinylidene fluoride impedance sensor forrespiratory event classification during polysomno-graphy. J Clin Sleep Med 2011; 7: 479–485.

55. Krauchi K, Cajochen C, Wirz-Justice A. Waking upproperly: is there a role of thermoregulation in sleepinertia? J Sleep Res 2004; 13: 121–27.

56. Dijk DJ, Boulos Z, Eastman CI, et al. Light treatmentfor sleep disorders: consensus report. II. Basicproperties of circadian physiology and sleep regu-lation. J Biol Rhythms 1995; 10: 113–125.

57. Hauri P. Evaluation of a sleep switch device. Sleep1999; 22: 1110–1117.

58. Sano A, Picard RW. Toward a taxonomy of autonomicsleep patterns with electrodermal activity. Conf ProcIEEE Eng Med Biol Soc 2011; 777–780.

59. Jones V, Bults R, de Wijk R, et al. Experience withusing the Sensewear BMS Sensor system in thecontext of a health and wellbeing application.Int J Telemed Appl 2011; ID; 671040: 1–18.

60. Sanchez JM, Pereira B, Paiva T. Headset Bluetoothand cell phone based continuous central bodytemperature measurement system. Conf Proc IEEEEng Med Biol Soc 2010; 2010: 2975–2978.

61. De Chazal P, Fox N, O’Hare E, et al. Sleep/wakemeasurement using a non-contact biomotion sensor.J Sleep Res 2011; 20: 356–366.

62. Zaffaroni A, de Chazal P, Heneghan C, et al.SleepMinder: an innovative contact-free device forthe estimation of the apnoea-hypopnoea index. ConfProc IEEE Eng Med Biol Soc 2009; 7091–7094.

63. Zaffaroni A, Kent B, O’Hare E, et al. Assessment ofsleep-disordered breathing using a non-contact bio-motion sensor. J Sleep Res 2013; 22: 231–236.

64. Murthy JN, van Jaarsveld J, Fei J, et al. Thermalinfrared imaging: a novel method to monitor airflowduring polysomnography. Sleep 2009; 32: 1521–1527.

65. Hers V, Corbugy D, Joslet I, et al. New concept usingPassive Infrared (PIR) technology for a contactlessdetection of breathing movement: a pilot studyinvolving a cohort of 169 adult patients. J Clin MonitComput 2013 [In press DOI: 10.1007/s10877-013-9457-2].

66. Takemura Y, Sato JY, Nakajima MN. A respiratorymovement monitoring system using fiber-gratingvision sensor for diagnosing sleep apnea syndrome.Opt Rev 2005; 12: 46–53.

67. Salmi T, Leinonen L. Automatic analysis of sleeprecords with static charge sensitive bed.Electroencephalogr Clin Neurophysiol 1986; 64: 84–87.

68. Tenhunen M, Elomaa E, Sistonen H, et al. Emfitmovement sensor in evaluating nocturnal breathing.Respir Physiol Neurobiol 2013; 187: 183–189.

69. Kogure T, Shirakawa S, Shimokawa M, et al.Automatic sleep/wake scoring form body motion inbed: validation of a newly developed sensor placedunder the mattress. J Physiol Anthropol 2011; 30:103–109.

70. Merilahti J, Saarinen A, Parkka J, et al. Long-termsubjective and objective sleep analysis of total sleepitme and sleep quality in real life settings. Conf ProcIEEE Eng Med Biol Soc 2007; 5202–5205.

71. Nishyama M, Miyamoto M, Watanabe K. Respirationand body movement analysis during sleep in bedsuing hetero-core fiber optic pressure sensorswithout constraint to human activity. J Biomed Opt2011; 16: 017002–1-7.

72. Willemen T, Van Deun D, Verhaert V, et al. Heartrate-, breathing rate- and movement-based sleep

Sleep medicine


classification. Proc Autumn Meeting Belgian AssocSleep Res Sleep Med 2012.

73. Zhu X, Chen W, Nemoto T, et al. Long-termmonitoring of heart rate, respiration rhythm, andbody movement during sleep based upon a network.Telemed J E-Health 2010; 244–252.

74. Shin JH, Chee YJ, Jeong DU, et al. Nonconstrainedsleep monitoring system and algorithms using air-mattress with balancing tube method. IEEE Trans InfTechnol Biomed 2010; 14: 147–156.

75. Wang F, Tanaka M, Chonan S. Development of aPVDF piezopolymer sensor for unconstrained in-sleep cardiorespiratory monitoring. J Intel Mat SystStr 2003; 14: 185–190.

76. Brink M, Muller CH, Schierz C. Contact-free meas-urement of heart rate, respiration rate, and bodymovements during sleep. Behav Res Methods 2006;38: 511–521.

77. Migliorini M, Bianchi AM, Nistico D, et al. Automaticsleep staging based on ballistocardiographic signalsrecorded through bed sensors. Conf Proc IEEE EngMed Biol Soc 2010; 3273–3276.

78. Beattie ZT, Hayes TL, Guilleminault C, et al. Accuratescoring of the apnea–hypopnea index using a simplenon-contact breathing sensor. J Sleep Res 2013; 22:356–362.

79. Di Rienzo M, Meriggi P, Rizzo F, et al. Textiletechnology for the vital signs monitoring in tele-medicine and extreme environments. IEEE Trans InfTechnol Biomed 2010; 14:3, 711–717.

80. Lymberis A, Dittmar A. Advanced wearable healthsystems and applications. IEEE Eng Med Biol Mag2007; 26: 29–33.

81. Agatsuma T, Fijmoto K, Komatsu Y, et al. A noveldevice (SD-101) with high accuracy for screeningsleep apnoea-hypopnoea syndrome. Respirology2009; 14: 1143–1150.

82. Ishida R, Yonezawa Y, Maki H, et al. A wearable,mobile phone-based respiration monitoring systemfor sleep apnea syndrome detection. Biomed SciInstrum 2005; 41: 289–293.

83. Smith CE, Dauz ER, Clements F, et al. Telehealthservices to improve nonadherence: a placebo-con-trolled study. Telemed J E-Health 2006; 12: 289–296.

84. Taylor Y, Eliasson A, Andrada T, et al. The role oftelemedicine in CPAP compliance for patients withobstructive sleep apnea syndrome. Sleep Breath2006; 10: 132–138.

85. Stepnowsky CJ Jr, Palau JJ, Marler MR, et al. Pilotrandomized trial of the effect of wireless telemoni-toring on compliance and treatment efficacy inobstructive sleep apnea. J Med Internet Res 2007;9: e14.

86. Kwiatkowska M, Idzikowski A, Matthews L.Telehealth-based framework for supporting thetreatment of obstructive sleep apnea. Stud HealthTechnol Inform 2009; 143: 478–483.

87. Cooper CB. Respiratory applications of telemedicine.Thorax 2009; 64: 189–191.

88. Farre R. The future of telemedicine in the manage-ment of sleep-related respiratory disorders. ArchBronchoneumol 2009; 45: 109–110.

89. Dellaca R, Montserrat JM, Govoni L, et al. TelemetricCPAP titration at home in patients with sleep apnea–hypopnea syndrome. Sleep Med 2011; 12: 153–157.

90. Pelletier-Fleury N, Lanoe JL, Philippe C, et al.Economic studies and ‘‘technical’’ evaluation oftelemedicine: the case of telemonitored polysomno-graphy. Health Policy 1999; 49: 179–194.

91. Gagnadoux F, Pelletier-Fleury N, Philippe C, et al.Home unattended vs hospital telemonitored poly-somnography in suspected obstructive sleep apneasyndrome: a randomized crossover trial. Chest 2002;121: 753–758.

92. Bruyneel M, Sanida C, Art G, et al. Sleep efficiencyduring sleep studies: results of a prospective studycomparing home-based and in-hospital based poly-somnography. J Sleep Res 2011; 20: 201–206.

93. Coma-del-Corral MJ, Alonso-Alvarez ML, Allende M,et al. Reliability of telemedicine in the diagnosis andtreatment of sleep apnea syndrome. Telemed J E-Health 2013; 19: 7–12.

94. Kristo DA, Eliasson AH, Poropatich RK, et al.Telemedicine in the sleep laboratory: feasibility andeconomic advantages of polysomnograms trans-ferred online. Telemed J E-Health 2001; 7: 219–224.

95. Seo J, Choi J, Choi B, et al. The development of anonintrusive home-based physiologic signal meas-urement system. Telemed J E-Health 2005; 11:487–495.

96. Choi JM, Choi BH, Seo JW, et al. A system forubiquitous health monitoring in the bedroom via aBluetooth network and wireless LAN. Engineering inMedicine and Biology Society 2004. EMBC 2004,conference proceedings. IEEE Proc 2004; 2:3362–3365.

97. Hsu YL, Yang CC, Tsai TC, et al. Development of adecentralized home telehealth monitoring system.Telemed J E-Health 2007; 13: 69–78.

98. Kayyali HA, Weimer S, Frederick C, et al. Remotelyattended home monitoring of sleep disorders.Telemed J E-Health 2008; 14: 371–374.

99. Alquassim S, Ganesh M, Khoja S, et al. 14th IEEEIntern Conf on E-health Networking, Application andServices (HealthCom), Beijing, China 2012; 443–446.

Sleep medicine


applications of evolving - european respiratory society · market moves apace with new...

Documents