F UL L ART I C L E

Assessing hemorrhagic shock: Feasibility of using anultracompact photoacoustic microscope

Qian Chen1,2 | Heng Guo1,2 | Weizhi Qi1,2 | Qi Gan3 | Lei Yang4 | Bowen Ke4 |

Xingxing Chen2 | Tian Jin1,2 | Lei Xi1,2*

1Department of Biomedical Engineering, SouthernUniversity of Science and Technology, Shenzhen,China2School of Physics, University of ElectronicScience and Technology of China, Chengdu,China3Department of Neurosurgery, West ChinaHospital Sichuan University, Chengdu, China4Department of Anesthesiology and Critical CareMedicine, West China Hospital SichuanUniversity, Chengdu, China

*CorrespondenceLei Xi, Department of Biomedical Engineering,Southern University of Science and Technology,Shenzhen, Guangdong, 518055, China.Email: xilei@sustc.edu.cn

Funding informationNational Natural Science Foundation of China,Grant/Award Numbers: 61775028, 81571722,61528401; Startup grant from Southern Universityof Science and Technology

Hemorrhagic shock, as an importantclinical issue, is regarding as a criticaldisease with a high mortality rate.Unfortunately, existing clinical tech-nologies are inaccessible to assess thehemorrhagic shock via hemodynamicsin microcirculation. Here, we proposean ultracompact photoacoustic micro-scope to assess hemorrhagic shock using a rat model and demonstrate its clinicalfeasibility by visualizing buccal microcirculation of healthy volunteers. Both func-tional and morphological features of the microvascular network including concen-tration of total hemoglobin (CHbT), number of blood vessels (VN), small vasculardensity (SVD) and vascular diameter (VD) were derived to assess the microvascu-lar hemodynamics of different organs. Animal studies show the feasibility of theproposed tool to assess and stage the hemorrhagic shock via microcirculation.in vivo oral imaging of healthy volunteers indicates the translational possibility ofthis technique for clinical evaluation of hemorrhagic shock.

1 | INTRODUCTION

Hemorrhagic shock remains an important clinical issue, whichmay lead to hemodynamic instability, decreases in oxygendelivery, insufficient tissue perfusion, cellular hypoxia, organdamage and even death [1–4]. All the clinical imaging tech-niques including X-ray computed tomography (CT), ultraso-nography and magnetic resonance imaging (MRI) are limitedto assess the hemorrhagic shock via hemodynamics in micro-circulation [5–7]. However, microcirculation, which is inac-cessible to clinical imaging modalities, plays a moreimportant role to assess and stage the hemorrhagic shock andfluid resuscitation [8]. With the rapid development of laserand light detection techniques, various optical imaging modal-ities including near-infrared spectroscopy (NIRS), pure opticalmicrocopy, orthogonal polarization spectral imaging, optical

intrinsic signal imaging and laser speckle contrast imaginghave been widely applied to study microcirculation of hemor-rhagic shock [9–11]. Among all these optical imaging modali-ties, NIRS suffers from low spatial resolution, fluorescencemicroscopy introduces the external contrast agents, both opti-cal intrinsic and laser speckle imaging lack of depth resolvingcapability. Apart from pure optical imaging modalities, opticalresolution photoacoustic microscopy (ORPAM), featuringrich optical contrast, deep penetration depth, multiscaleresolving capability and ultrahigh sensitivity to hemoglobin,shows significant advantages on visualizing microcirculationof rodents [12–14]. Unfortunately, clinical translation ofORPAM is still challenging due to the bulky size and incon-venience of the equipment. Recent clinical investigationsshow that the microcirculatory indices in sublingual or buccalmucosa reflect the systemic microcirculation which is tightlyassociated with hemorrhagic shock [8]. Our group proposed arotatory-scanning-based portable ORPAM and reported theQian Chen and Heng Guo contributed equally to this work

Received: 13 September 2018 Revised: 27 October 2018 Accepted: 9 November 2018

DOI: 10.1002/jbio.201800348

J. Biophotonics. 2018;e201800348. www.biophotonics-journal.org © 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1 of 9https://doi.org/10.1002/jbio.201800348

first study of human buccal microcirculation [15]. Although itis portable relative to tabletop ORPAMs and achieves high-resolution imaging of buccal microcirculation in humanrevealing the potential for clinical use, it is still heavy andbulky for handheld devices and difficult to hold stably for along time. To further reduce the portable probe and make itsuitable for clinical use, we utilized a two-dimensional(2D) electrothermal-bimorph-actuation based micro-electro-mechanical system (MEMS) scanner to develop a miniaturizedORPAM system, a handheld microscope and an ultracompactORPAM (U-ORPAM) probe [16–19]. This probe weights20 g and has an outer size of 22 × 30 × 13 mm3, a high lat-eral resolution of 3.8 μm, an effective imaging domain of2 × 2 mm2 and a maximal full-view frame rate of 2 Hz.

In this study, our purpose is to investigate the microvas-cular hemodynamics of different organs during the entirecourse of hemorrhagic shock and resuscitation with the U-ORPAM, which has been successfully applied to monitorhemodynamics in human buccal vasculatures.

2 | MATERIALS AND METHODS

2.1 | Animal preparation

The experimental protocol was approved by the AnimalEthics Committee at the Southern University of Science andTechnology (SUSTech). Female Sprague-Dawley rats,weighting 250 to 300 g (n = 60), were equally divided intofour groups. The rats were anesthetized by intraperitonealinjection of chloral hydrate (50 mg/kg). Anesthesia wasmaintained by additional supplements of chloral hydrate atintervals of approximately 30 to 45 minutes. We carried outexperiments of four organs including brain, intestine, ovaryand bladder. The rats in each group were evenly divided intothree sub-groups:(a) shock group: the rats received surgeries,induction of hemorrhagic shock and blood resuscitation;(b) control group: the rats undertook surgeries withoutinducing shock; (c) death group: the rats received surgeriesas well as the induction of shock without resuscitation.

2.2 | The shock models

We established two rat shock models including the standardone and the rapid one. We evaluated the cerebral hemody-namics with standard shock model. However, considering thatthe internal organs cannot be exposed to the air for a longtime, we only carried out experiments for internal organsbased on the rapid shock model. For both models, Two PE-50cannulas were inserted into the femoral arteries on both sides.One cannula was connected to a pressure transducer to moni-tor the mean arterial pressure using a multi-parametric physio-logical recording device (RM6240, Chengdu Instrument Inc.,Chengdu, China). We extracted blood from the other cannulato induce hemorrhagic shock. One more cannula was inserted

into the femoral vein on the right side to infuse blood mixedwith heparin sodium solution. A feedback-controlled heatingpad was used to maintain the body temperature at37.0 � 0.5�C over the entire course of the experiment. Forbrain imaging, we did craniotomies to remove both the scalpand skull. The dura was kept intact and continually bathedwith artificial cerebrospinal fluid. For experiments of abdomi-nal organs, we expose the organ with minimal invasive inci-sion to avoid unnecessary exposure of other organs.

For the experiments with the standard shock model, wecarried out full-view imaging of the brain with an interval of2 minutes within 10 minutes prior to the extraction of blood.Then, the blood was extracted with a constant speed from thefemoral vein using a syringe pump to target a mean arterialpressure of 40 mm Hg within 15 minutes. We captured10 more images during bleeding. Hypovolemia was maintainedfor additional 1 hour and we performed 18 experiments tomonitor the microvascular hemodynamics. In the stage ofresuscitation, the blood was reinfused via femoral vein andadditional six images were captured within 20 minutes. Wespent another 80 minutes to monitor the rat with 24 experimentsto monitor the microvascular hemodynamics post resuscitation.

For the rapid model, we firstly carried out five full-viewimaging experiments as the resting state within 5 minutes,then extracted the blood with a constant velocity and contin-ued to image the microvascular hemodynamics every minuteuntil the arterial pressure dropped to 40 mm Hg withinanother 5 minutes. We imaged the same region (15 images)when the arterial pressure was maintained around 40 mmHg for additional 15 minutes. After that, the blood resuscita-tion (5 images) was carried out within 5 minutes. Post theresuscitation, we carried out 15 experiments to monitor themicrovascular hemodynamics until the arterial pressurereturn to a stable level.

2.3 | The imaging system

Supplementary Figure S1, Supporting Information presentsthe configuration of the system, which is detailed describedin Supplementary Section I. Figure 1A shows the photo-graph of the system integrating all the optical and electricalcomponents inside a movable cart. The probe highlighted inFigure 1B has an outer size of 22 × 30 × 13 mm3, a weightof 20 g, a lateral resolution of 3.8 μm (SupplementaryFigure S1B), an axial resolution of 104 μm (SupplementaryFigure 1C) and a frame rate of up to 2 Hz with a frame sizeof 500 × 500 pixels in a 2 × 2 mm2 imaging domain.

2.4 | Image reconstruction

The raw photoacoustic signals were filtered using a bandpassfilter (5-20 MHz) to remove both low-frequency and high-frequency electromagnetic noise. All depth-resolved A-lineswere processed using a Hilbert transform and directly back-projected to a rectangular coordinate. We showed all

2 of 9 CHEN ET AL.

volumetric data through projecting the maximal amplitudeof each photoacoustic signal (MAP). The movie was per-formed using Matlab (MathWorks Inc., Natick, Massachu-setts) by stacking MAP images captured at different timepoints. We derived the concentration of total hemoglobin(CHbT), number of blood vessels (VN), vascular density ofsmall blood vessels (SVD) and vascular diameter (VD) toassess the hemorrhagic shock and associated blood resuscita-tion using a home-developed algorithm described in Supple-mentary Section II.

2.5 | Statistical analysis

We carried out the graphic display of data and statisticalanalysis using Graphpad Prism 6 (San Diego, California),and reported the values as mean � standard deviation (SD).Kolmogorov-Smironv test was used to confirm normal dis-tribution of experimental data. The parametric test (unpairedt test) and nonparametric test (Mann-Whitney U test) werecarried out to do comparison between time-based experi-mental data within each group. Two-way ANOVA

(Tukey's multiple comparisons test) was performed betweenshock-normal, shock-death groups for each organ to com-pare relative microvascular hemodynamics. We considerP < 0.05 as statistically significant.

2.6 | The human experiments

We imaged the tongues and lips of three male volunteers.All volunteers wore the protection glass to avoid potentiallaser exposure to the eyes. The lips and tongues wereattached to the imaging interface via drinkable water. Afterthe experiments, dentists continued to examine the imagedarea for 7 days and no abnormal symptom was observed.We have obtained the written consent forms from all the vol-unteers participating in the experiments.

3 | RESULTS

Figure 1C shows a photograph of a volunteer participatingin the human experiment, in which an operator holds the

FIGURE 1 System configurations and in vivo human oral imaging. A, A photograph of the integrated system consisting of data acquisition system,electrical and optical components. B, The schematic of the imaging probe marked by the red dash box in (A). The ultracompact probe has an outer size of22 × 30 × 13 mm3 and a weight of 20 g. C, A healthy volunteer participates in the experiment of oral imaging. D and E, The maximum amplitude projection(MAP) of the buccal microvasculature on the up lip and hypogloeeis of the volunteer. The effective imaging domain is 2 × 2 mm2. CL, collimation lens; DL,doublet lens; P, right-angle prism; MEMS, micro-electro-mechanical system; CG, cover glass; T, transducer

CHEN ET AL. 3 of 9

probe and images the region of interest in the oral cavity.Figure 1D,E present the MAP images of microvasculaturelocated on the superficial surfaces of the upper lip and hypo-gloeeis. A large number of blood vessels forming an ultra-dense microvascular network are clearly visualized. On thetop surface of the buccal mucosa, there exists many capillaryloops which change at the beginning of hemorrhagic shock.

Figure 2 shows the MAP images of a typical rat brainobtained at different stages of standard hemorrhagic shock.Figure 2A shows the photoacoustic image of cerebral cortexprior to bleeding, referred to as the baseline. Figure 2B,Cpresent the cerebral hemodynamics of the rat with the loss of20% and 35% of the total blood volume, respectively. Due toinsufficient supply of blood, CHbT declines to 55% and 35%of the baseline (Figure 3), respectively. Post bleeding, asshown in Figures 2D,E and 3, CHbT slightly recovers to 60%of the baseline due to the compensated perfusion by auto-regulation mechanism of the body and then declines to 35%of the normal status because of the failure in compensatorymechanism. The close-up views of sub-region indicated bywhite dashed boxes in Figure 2C-E are presented in the bot-tom of Figure 2, marked by I, II and III. The quantitatedphotoacoustic intensities of a selected blood vessel revealthe compensatory mechanism. Figure 2F-H shows the recov-ery of the microcirculation during and post the procedure ofblood resuscitation. Owning to the blood infusion, CHbT

gradually return to the baseline. Figure 2I shows the changeof the same blood vessel, marked by blue, green and reddashes, respectively. We can see that there is limited changeof vessel diameter, but significant change of photoacousticamplitude after bleeding.

We notice that VN, SVD and VD changed mildly whilethe CHbT has a more intense change over the entire course ofhemorrhagic shock and blood resuscitation (Figure 3). VN,SVD and CHbT show a consistent change tendency withinthe entire course of hemorrhagic shock, while the recoveryof vessel diameter is not obviously. It is possible that bloodredistribution and vasodilatation, which change the diameterof blood vessels, is much more complicated than VN, SVDand CHbT. Visualization 1 depicts the microvascular kineticsover the entire period of the hemorrhagic shock andresuscitation.

Figure 4 presents the hemodynamics in the microcir-culation of intestine, ovary and bladder. Visualization2 shows the longitudinal monitoring of the microvascularhemodynamics of a selected intestine, ovary and bladderover the entire course of experiments. We selected fiveimages at the stages of baseline, bleeding, shock, bloodinfusion and postinfusion. We observe more intensechanges of microvascular morphology compared with thatof cerebral cortex. For example, with the loss of blood, alarge amount of small blood vessels disappear. In

FIGURE 2 The representative photoacoustic MAP images of brain vasculature at different time points over the entire course of hemorrhagic shock andblood resuscitation including the resting state (A), loss of 20% of the total blood volume (B), loss of 35% of the total blood volume (C), compensatory stage(D), decompensatory stage (E), blood infusion with 70% of the total blood volume (F), fully blood infusion (G), 1 hour after blood resuscitation (H). Figures I,II and III show the close-up views of the regions indicated by the dash white box in (C), (D) and (E), respectively; I, Profiles along the dashed lines in (I),(II) and (III), respectively. The PA intensity slightly increases because of the compensatory mechanism and declines due to the failure of compensatorymechanism

4 of 9 CHEN ET AL.

FIGURE 3 Quantitative analysis of functional and morphological features of the cerebral cortex over the entire course of hemorrhagic shock. Relativechanges of vascular number (A), small vascular density (B), total hemoglobin concentration (C) and vessel diameter (D). We normalized all the data for eachparameter and show the data at stages of baseline (0-10 minutes), rapid bleeding (10-25 minutes), compensatory (25-35 minutes), decompensatory(35-85 minutes), rapid blood infusion (85-105 minutes) and postresuscitation (105-185 minutes). P < 0.05: *; P < 0.01: **; P < 0.001: ***;P < 0.0001: ****

FIGURE 4 Microvascular hemodynamics of internal organs during hemorrhagic shock and blood resuscitation. A, Five selected MAP images of an intestineat stages of baseline, bleeding, shock, infusion and post infusion. B, Five selected MAP images of an ovary at stages of baseline, bleeding, shock, infusionand post infusion. C, Five selected MAP images of a bladder at stages of baseline, bleeding, shock, infusion and post infusion. Yellow rows indicate typicalvessels in different organs, which shrink with the loos of blood and recover post the resuscitation. Scale bar: 400 μm

CHEN ET AL. 5 of 9

addition, the vasoconstriction of large vessels indicated bythe yellow arrows is significant. Furthermore, we find thatthe loss of blood will result in the shrink of the organs andfurther lead to the slight motion of the imaging domain atdifferent stages. It is worth to notice that, organs could notfully recover after the blood perfusion within a short time.Apart from the microvascular hemodynamics in theabdominal organs, there is no loss of the small vessels andlimited morphological change of large and medium ves-sels in the cerebral cortex during the entire course ofbleeding and infusion.

Figure 5 shows the microvascular hemodynamics of con-trol and death groups. In the control group, there is no obvi-ous change of microcirculation. Visualization 3 shows themicrovascular hemodynamics of a representative intestine,ovary and bladder during the experiments in the controlgroups, respectively. In contrast, there is no recovery of vas-cular hemodynamics in the death group. Visualization 4 pre-sents the longitudinal data of a representative intestine,ovary and bladder in the death groups, respectively.

Figure 6 shows the quantitative analysis of CHbT, VN,SVD and VD of the abdominal organs in different groups. In

FIGURE 5 The photoacoustic images of internal organs in control groups. A, C, E, The microvascular hemodynamics of intestine, ovary and bladder in thegroup without bleeding. B, D, F, The microvascular hemodynamics of intestine, ovary and bladder in the group without resuscitation. For all the controlcases, we collected data for 45 minutes and selected MAP images corresponding to the cases in the shock group

6 of 9 CHEN ET AL.

control groups, all the derived parameters slightly declinedue to the long-time exposure of organs to the air, and donot return to the base line after the infusion of blood. Theparameters of both shock and death groups have the similarvariation tendency at the stages of baseline, bleeding andshock. As we expected, the functions of vasculature partiallyreturn to normal status after the blood resuscitation in shockgroups, while, the organs of death groups fail to recover dueto the severe loss of blood. Supplementary Tables S1 and S2summarize the statistical analysis of the data in shock-control groups and shock-death groups.

4 | DISCUSSION

Generally, a typical hemorrhagic shock includes three majorstages: compensatory-, decompensatory- and refractory-stage [8, 11, 20]. In compensatory-stage, abnormal bleedingtriggers the auto-regulatory mechanism of the body, whichredistributes the blood in microvasculature and peripheralorgans to maintain the minimum perfusion of vital organs.In decompensatory-stage, the microvascular congestion oforgans leads to the failure of auto-regulatory mechanism and

potential damage to vital organs. In the last stage featuringrefractory hypotension and failure of microcirculation, thebody will suffer from irreversible cell damage, organ disabil-ity and even death. The cerebral data clearly presents thesuddenly rise in CHbT after the termination of bleeding,which represents the occurrence of compensating mecha-nism of the body. In addition, we observe slight vasocon-striction of vessels, which possibly aims to moderate thesystemic arterial pressure by increasing systemic vascularresistance [21]. However, we note that the vessel numberand small vascular density do not variate dramatically duringthe entire course of hemorrhagic shock and blood resuscita-tion. Evidence indicating that the cerebral flow is redirectedfrom peripheral organs to the brain has been documented[22]. Furthermore, the unique response of cerebral microcir-culation to hemorrhagic shock, in which microcirculatoryflow is sustained at essentially normal levels, has beenreported [8, 11, 23]. Benefiting from the principle of photoa-coustic effect which is ultrasensitive and positively propor-tional to the concentration of total hemoglobin at 532 nm,we clearly observe significantly decline of CHbT at the stagesof bleeding and shock. Even the auto-regulatory mechanismpartially compensates the decline of CHbT and protects the

FIGURE 6 Quantitative and statistical analysis of CHbT, VN, SVD and VD of microvascular networks in intestine, ovary and bladder in shock, control anddeath groups. The left, middle, right columns show the results of intestine, ovary and bladder, respectively. We normalized the data for each parameter anddisplay the data using mean � SD

CHEN ET AL. 7 of 9

cerebral tissue from ischemic injury, CHbT will eventuallydeclines due to the insufficient supply of blood.

Apart from cerebral microcirculations, abdominal organsincluding intestine, ovary and bladder, which are less essentialthan brain, have much more significant microvascular hemo-dynamics during the entire course of hemorrhagic shock andresuscitation. Similar to cerebral microcirculation, CHbT

declines with bleeding and recovers with infusion. However,we observe the disappearance of small blood vessels, espe-cially capillaries, within the stages of bleeding and shock,which is totally different from the cerebral hemodynamics. Inaddition, the vasoconstriction of the vessels in abdominalorgans is more obvious than that of cerebral cortex. Interest-ingly, we notice that intestine and ovary are capable of fullyrecovering, while bladder cannot return to the normal statuseven at several hours post the blood infusion. The major rea-sons may include: (a) the importance of abdominal organs isdifferent resulting in the diverse response to resuscitation;(b) the response speed of each organ is variated.

In human experiments, we observe ultra-dense vascularnetworks formed by large, medium and small vessels aswell as capillary loops which connects the venule and arte-riole, will change at the early stage of bleeding and istightly associated with the severity of hemorrhagic shock.Successful visualization of human buccal microcirculationin three-dimensional with capillary-level resolving capabil-ity reveals the great clinical translational potential of thistechnique.

Although our study is encouraging, we still recognizesome important limitations in the experimental methodand the interpretation of our findings. Firstly, the use ofsingle wavelength illumination only enables the derivationof limited parameters which are insufficient to evaluatehemorrhagic shock in clinic. However, this is not the fun-damental issue of photoacoustic. Through the utilizationof multispectral strategy and acoustic Doppler effect, it isfeasible to estimate the concentrations of oxygen- anddioxygen- hemoglobin, oxygen saturation and blood flow,making the assessment more accurate [24, 25]. Secondly,the imaging depth is still limited by both the transparencyof tissue and excitation wavelength. With tissue clearingtechniques [26] and near-infrared illumination [27], wewill observe more blood vessels in deep locations.Thirdly, it is hard to explain how the responses of periph-eral organs to hemorrhagic shock or ischemia are differentbased on current data. In further studies, other opticalimaging techniques such as optical coherence tomography(OCT), laser speckle imaging et al. should be combinedwith the proposed techniques to derive more functional,morphological and molecular parameters for furtherassessment of organ response to hemorrhagic shock andfluid resuscitation. Last, it is feasible to increase the imag-ing speed by using a faster scanner, making it possible to

capture fast events of microcirculations during hemor-rhagic shock.

ACKNOWLEDGMENTS

This work was sponsored by National Natural Science Foun-dation of China (61775028, 81571722 and 61528401),Startup grant from Southern University of Science and Tech-nology. The authors thank the volunteers for participating inthe human experiments.

Author contributions

L.X. conceived the concept, proposed the experimentalplans, supported and supervised the project. L.X., Q.C.,H.G., X.C. and T.J. built the system, acquired and processedthe data. W.Q. and H.G. provided the assistance in opera-tion, assembly of the MEMS scanner. Q.G., L.Y. andB.K. assisted the establishment of the disease model. L.X.,Q.C. and W.Q. prepared the manuscript.

ORCID

Qian Chen https://orcid.org/0000-0002-0815-5336

Lei Xi https://orcid.org/0000-0002-2598-6801

REFERENCES

[1] K. K. Nandra, M. Collino, M. Rogazzo, R. Fantozzi, N. S. Patel,C. Thiemermann, Dis. Model Mech. 2013, 6, 701. https://doi.org/10.1242/dmm.011353.

[2] D. Wu, X. Zhou, Y. Liganf, J. Gan, M. Zhang, Acad. Emerg. Med. 2015,22, 915.

[3] X. Dunquan, G. Cao, N. Wen, L. Yan, W. Yanxia, G. Changjun, D. Qian,Y. Linong, C. Wei, L. Zhichao, Acta Pharmacol. Sin. 2013, 34, 1515.https://doi.org/10.1038/aps.2013.96.

[4] E. E. Douzinas, Crit. Care Med. 2012, 40, 1348. https://doi.org/10.1097/CCM.0b013e31823e9501.

[5] C. J. Sivit, G. A. Taylor, D. I. Bulas, D. C. Kushner, B. M. Ppotter,M. R. Eichelberger, Radiology 1992, 182, 723. https://doi.org/10.1148/radiology.182.3.1535886.

[6] F. Corradi, C. Brusasco, A. Vezzani, S. Palermo, F. Atomonte,P. Moscatelli, P. Pelosi, Radiology 2011, 260, 112. https://doi.org/10.1148/radiol.11102338.

[7] K. C. Li, L. R. Pelc, S. Puvvala, G. A. Wright, Radiology 1998, 206, 219.https://doi.org/10.1148/radiology.206.1.9423676.

[8] W. Zhi, S. Shijie, G. Ristagno, M. H. Weil, T. Wanchun, Crit. Care Med.2010, 38, 928. https://doi.org/10.1097/CCM.0b013e3181cd100c.

[9] G. J. Beiman, K. E. Groehler, V. Lazaron, J. P. Ortner, Shock 1999, 12, 196.[10] Z. Rong, D. Xiaoli, L. Liangming, Acta Pharmacol. Sin. 2014, 35, 1375.

https://doi.org/10.1038/aps.2014.83.[11] S. Nannan, L. Weihua, L. LinZ, L. Qingming, Acad. Radiol. 2014, 21, 175.

https://doi.org/10.1016/j.acra.2013.11.017.[12] T. T. W. Wong, R. Zhang, P. Hai, C. Zhang, M. A. Pleitez, R. L. Aft,

D. V. Novack, L. V. Wang, Sci. Adv. 2017, 3, e1602168. https://doi.org/10.1126/sciadv.1602168.

[13] P. Manivasagan, N. Q. Bui, S. Bharathiraja, M. S. Moorthy, O. Yun-Ok,K. Song, S. Hansu, M. Yoon, O. Junghwan, Sci. Rep. 2017, 7, 43593.https://doi.org/10.1038/srep43593.

[14] X. Zhixing, W. Roberts, P. Carson, L. Xiaojun, T. Chao, W. Xueding, Opt.Lett. 2011, 36, 4815. https://doi.org/10.1364/OL.36.004815.

[15] J. Tian, G. Heng, Y. Lei, X. Huikai, J. Huabei, X. Lei, J. Biophotonics2018, 11, e201700250. https://doi.org/10.1002/jbio.201700250.

8 of 9 CHEN ET AL.

[16] C. Qian, G. Heng, J. Tian, Q. Weizhi, X. Huikai, X. Lei, Opt. Lett. 2018,43, 1615. https://doi.org/10.1364/OL.43.001615.

[17] W. Z. Qi, Q. Chen, H. Guo, H. K. Xie, L. Xi,Micromachines 2018, 9(288), 1.[18] W. Qin, Q. Chen, L. Xi, Biomed. Opt. Express 2018, 9(5), 2205.[19] H. Guo, Q. Chen, W. Z. Qi, X. X. Chen, L. Xi, J. Biophotonics 2018, 11,

e201800067. https://doi.org/10.1002/jbio.201800067.[20] G. Gutierrez, H. D. Reines, W.-G. ME, Crit. Care 2004, 8, 373. https://doi.

org/10.1186/cc2851.[21] P. M. Reilly, K. B. Wilkins, K. C. Fuh, Shock 2001, 15, 329.[22] G. Slater, B. C. Vladeck, R. Bassin, W. C. Shoemaker, Ann. Surg. 1975,

181, 1. https://doi.org/10.1097/00000658-197501000-00001.[23] D. D. Heistad, Microcirculation 2001, 8, 365.[24] J. M. Yang, C. Favazza, C. Ruimin, Y. Junjie, C. Xin, K. Maslov, Z. Qifa,

K. K. Shung, W. Lihong, Nat. Med. 2012, 18, 1927. https://doi.org/10.1038/nm.2823.

[25] K. S. Valluru, K. E. Wilson, J. K. Willmann, Radiology 2016, 280, 332.https://doi.org/10.1148/radiol.16151414.

[26] Z. Yanjie, Y. Tingting, Z. Chao, L. Zhao, L. Qingming, X. Tonghui,Z. Dan, Light Sci. Appl. 2018, 7, 17153. https://doi.org/10.1038/s41377-018-0021-1.

[27] Y. Junjie, A. A. Kaberniuk, L. Lei, D. M. Shcherbakova, Z. Ruiying,W. Lidai, L. Guo, V. V. Verkhusha, W. LihongV, Nat. Methods 2016, 13,67. https://doi.org/10.1038/nmeth.3656.

SUPPORTING INFORMATION

Additional supporting information may be found online inthe Supporting Information section at the end of the article.Figure S1 The schematic of the system configurations andthe ultracompact imaging probe (A). Estimation of the lateral(B) and axial (C) resolutions of the system. PC, personalcomputer; OL, objective; DAQ, data acquisition card; Amp,amplifier; PH, pinhole; L, convex lens; SMF, single modefiber; CL, collimation lens; DL, doublet lens; P, right-angle

prism; MEMS, micro-electro-mechanical system; GC, glasscover; FG, function generator (motor control); T, transducer.Figure S2 The flowchart of imaging postprocessing.Table S1 Statistical analysis of derived parameters betweenshock and control groups.Table S2 Statistical analysis of derived parameters betweenshock and death groups.

Video S1 The longitudinal monitoring of rat cerebral cortexand recording of mean arterial pressure during the entire courseof hemorrhagic shock and blood resuscitation within 3 hours.

Video S2 The longitudinal monitoring of three representa-tive internal organs and recording of mean arterial pressureduring the entire course of hemorrhagic shock and bloodresuscitation within 45 minutes.

Video S3 The longitudinal monitoring of three representa-tive internal organs and recording of mean arterial pressurein the control group within 45 minutes.

Video S4 The longitudinal monitoring of three representa-tive internal organs and recording of mean arterial pressurein the death group within 45 minutes.

How to cite this article: Chen Q, Guo H, Qi W,et al. Assessing hemorrhagic shock: Feasibility ofusing an ultracompact photoacoustic microscope. J.Biophotonics. 2018;e201800348. https://doi.org/10.1002/jbio.201800348

CHEN ET AL. 9 of 9