Intraoperative cerebral blood flow imaging of rodentsHangdao Li, Yao Li, Lu Yuan, Caihong Wu, Hongyang Lu, and Shanbao Tong

Citation: Review of Scientific Instruments 85, 094301 (2014); doi: 10.1063/1.4895657 View online: http://dx.doi.org/10.1063/1.4895657 View Table of Contents: http://scitation.aip.org/content/aip/journal/rsi/85/9?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Real-time full field laser Doppler imaging AIP Conf. Proc. 1457, 282 (2012); 10.1063/1.4730568 Robust surface registration using salient anatomical features for image-guided liver surgery: Algorithm andvalidation Med. Phys. 35, 2528 (2008); 10.1118/1.2911920 Laser range scanning for image-guided neurosurgery: Investigation of image-to-physical space registrations Med. Phys. 35, 1593 (2008); 10.1118/1.2870216 Noninvasive monitoring of traumatic brain injury and post-traumatic rehabilitation with laser-inducedphotoacoustic imaging Appl. Phys. Lett. 90, 243902 (2007); 10.1063/1.2749185 Dynamic, three-dimensional optical tracking of an ablative laser beam Med. Phys. 32, 209 (2005); 10.1118/1.1828672

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

202.120.62.150 On: Sat, 20 Sep 2014 01:50:49

REVIEW OF SCIENTIFIC INSTRUMENTS 85, 094301 (2014)

Intraoperative cerebral blood flow imaging of rodentsHangdao Li,1 Yao Li,1,2 Lu Yuan,1 Caihong Wu,3 Hongyang Lu,1 and Shanbao Tong1,2,a)

1Med-X Research Institute, Shanghai Jiao Tong University, Shanghai 200030, China2School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China3School of Media and Design, Shanghai Jiao Tong University, Shanghai 200240, China

(Received 20 June 2014; accepted 1 September 2014; published online 19 September 2014)

Intraoperative monitoring of cerebral blood flow (CBF) is of interest to neuroscience researchers,which offers the assessment of hemodynamic responses throughout the process of neurosurgery andprovides an early biomarker for surgical guidance. However, intraoperative CBF imaging has beenchallenging due to animal’s motion and position change during the surgery. In this paper, we pre-sented a design of an operation bench integrated with laser speckle contrast imager which enablesmonitoring of the CBF intraoperatively. With a specially designed stereotaxic frame and imager, wewere able to monitor the CBF changes in both hemispheres during the rodent surgery. The rotatabledesign of the operation plate and implementation of online image registration allow the technicianto move the animal without disturbing the CBF imaging during surgery. The performance of thesystem was tested by middle cerebral artery occlusion model of rats. © 2014 AIP Publishing LLC.[http://dx.doi.org/10.1063/1.4895657]

I. INTRODUCTION

Real time cerebral blood flow (CBF) serves as an impor-tant physiologic marker, which promptly reflects the in vivohemodynamic and cerebral circulation status. In rodent strokemodels, such as intraluminal filament model of middle cere-bral artery occlusion (MCAO),1 intraoperative CBF could beused to confirm the success of occlusion. It also provides theearly prediction of brain injury in MCAO and potential realtime guidance for the surgery operation to improve the model-ing stability.2 Moreover, continuous monitoring of CBF helpsto unravel mechanisms of chronic cerebral hypoperfusion tothe cognitive impairment, and thus optimize the time windowfor therapeutics.3, 4 Intraoperative CBF monitoring also hasa variety of clinical applications, e.g., the detection of cere-bral hypoperfusion in carotid endarterectomy,5 identificationof clip-related blood vessel stenosis or occlusion,6 evalua-tion of bypass patency during extracranial-intracranial bypassgrafts,7 and assessment of repurfusion of revascularization,8

etc.There have been various imaging techniques for moni-

toring CBF, including magnetic resonance imaging (MRI),9

positron emission tomography (PET),10 and cerebral perfu-sion computed tomography (CT),11 etc. However, the accep-tance of these techniques in intraoperative CBF monitoringwas hampered due to many factors, such as high cost, poortemporal or spatial resolution, use of radioactive tracers, andassessibility of the facilities.

Currently, the most used intraoperative CBF monitor-ing technique is laser Doppler flowmetry (LDF), which pro-vides continuous, noninvasive, and real time measurement ofblood flow in vessels.12, 13 However, the accuracy of LDFis compromised by the physical motion of the subject orthe orientation of the Doppler probe. Moreover, the sin-

a)Author to whom correspondence should be addressed. Electronic mail:stong@sjtu.edu.cn

gle point measurement might not be representative of over-all blood flow because of the spatial heterogeneity of themicrovasculature.14

Other optical methods can also be used for obtainingCBF information in vivo. Doppler optical coherence tomog-raphy (OCT)15 has great spatial resolution, which enables usto monitor capillary CBF. Photoacoustic tomography (PAT)16

is capable of imaging deep vessels for its strength in depthpenetration. However, both methods suffer from poor tempo-ral resolution due to time-consuming scanning, thus are notsuitable for real-time and intraoperative imaging.

In contrast, laser speckle contrast imaging (LSCI) isa non-contact full-field technique which provides two-dimension CBF information with high spatial and tempo-ral resolution.17 LSCI has been successfully used to get thestructural and functional changes of CBF in a wide rangeof biomedical applications.18–22 Clinically, CBF data couldbe obtained intraoperatively using LSCI in patients undergo-ing surgical revascularization.8, 23 It is noted that nearly allLSCI applications require the subjects to be placed in proneposition,8, 24, 25 while many neurosurgeries require the sub-jects to be placed in supine position,26, 27 thus making theexisting LSCI systems unsuitable for intraoperative imaging.Therefore, an intraoperative LSCI system is in great need toperform real time full field CBF imaging in a wide range ofbomedical studies.

In this work, we present a design of a novel operationsystem integrated with a LSCI imager for intraopreative CBFmonitoring during the experiment of rodents modeling. Thesystem consists of an operation bench specially designed foranimals in either supine or prone positions. The system is ro-bust to the motions due to surgical operations, which is par-ticularly useful in experimental studies, e.g., rodent MCAOmodel.28

We tested the system performance in rat model of MCAOexperiment. MCAO comprises a battery of vascular surg-eries and operations, which is similar to many clinical or

0034-6748/2014/85(9)/094301/7/$30.00 © 2014 AIP Publishing LLC85, 094301-1

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

202.120.62.150 On: Sat, 20 Sep 2014 01:50:49

094301-2 Li et al. Rev. Sci. Instrum. 85, 094301 (2014)

pre-clinical neurovascular surgeries. Besides, each phase ofMCAO surgery could lead to immediate cerebral blood flowchange, which makes it an ideal application to test the perfor-mance of the system.

II. MATERIALS AND METHODS

A. System design

The whole system consisted of a specially designed op-eration bench, a LSCI imager, and the graphic-user interface(GUI) software.

1. Operation bench

Fig. 1 shows the schematic of operation bench includ-ing four major components: (1) a workbench as the support-ing frame and workspace; (2) a round operation plate (OP),which can be rotated continuously up to 180◦ so that animalscan be moved intraoperatively to the appropriate surgical ori-entation. The OP is specially designed with a center window(2 cm in diameter) for CBF imaging in supine position. (3) Aspecially designed two-mode stereotaxic frame that can be in-

stalled with either a standard prone adaptor or a self-designedsupine adaptor (Fig. 1(a)). The stereotaxic frame was remod-eled from a commercially available setup (68001-Single Ma-nipulator Stereotaxic, RWD Life Science, Shenzhen, China).In order to fix the animal in supine position, the stereotaxicframe was re-designed so that the ear bar holders could beadjusted in z-direction (see the inset of zoom-in design inFig. 1). (4) An imager container attached to the OP fromthe bottom. When the technician moves the OP during thesurgery, the imager will move simultaneously so that thereal time CBF monitoring would not be interrupted. Fig. 1shows the 3D details of the major modules. Fig. 2 shows theschematics and the optical paths of the system working in bothsupine and prone positions. Fig. 3 shows the snapshots of thewhole system in an experiment when the rat is fixed in ei-ther supine position (Figs. 3(a) and 3(c)) or prone position(Figs. 3(b) and 3(d)).

2. LSCI imager

The overall design of the high spatiotemporal resolutionLSCI imager is shown in Fig. 1(d) (20 fps, 1280 × 1024pixels). The spatial and temporal resolution is 4.65 μm and

FIG. 1. 3D schematic of the system for intraoperatively monitoring cerebral blood flow. (a) and (b) Remodeled stereotaxic frame for fixing animal in supineor prone position; (c) rotatable plate as the operation area, enabling the adjustment of position during CBF monitoring and surgery; (d) and (e) the imager canmonitor the dynamic CBF changes from bottom (supine position) or upside (prone position) of the operation bench.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

202.120.62.150 On: Sat, 20 Sep 2014 01:50:49

094301-3 Li et al. Rev. Sci. Instrum. 85, 094301 (2014)

FIG. 2. Schematics and optical paths of the system in both (a) supine and (b) prone positions. The arrows indicate the optical paths.

FIG. 3. The real system in an experiment. (a) The real-time monitoring of CBF when the rat was in supine position. (b) The imaging of CBF of the rat in proneposition, suggesting the easy switch between two different modes. (c) and (d) Details of supine and prone fixation of the rat, respectively.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

202.120.62.150 On: Sat, 20 Sep 2014 01:50:49

094301-4 Li et al. Rev. Sci. Instrum. 85, 094301 (2014)

50 ms, respectively. And the imager is mounted on the top ofthe bench by a desk-mountable arm when working in proneposition as shown in Figs. 2(b) and 3(b). The schematic ofthe imager is illustrated in Fig. 1(e). The motion artifacts dueto respiration and surgical motion were suppressed by regis-tered laser speckle contrast analysis (rLASCA)29 algorithm.Nevertheless, the real time registration is time consuming. Toimprove the computational efficiency, we implemented thereal-time registration on NVIDIA’s parallel computing ar-chitecture (i.e., Compute Unified Device Architecture, orCUDA) based on graphics processing unit (GPU). As illus-trated in Fig. 4(c), the selected vessels drifted out of the re-gion of interest (ROI) during the surgery, which could be suc-cessfully corrected by online registration (see Figs. 4(d) and 5(Multimedia view)).

3. GUI-software

The imager was neatly connected to the computer via aUSB connection, plus a cable for power supply. Based on Mi-crosoft Foundation Classes (MFC, Microsoft Inc., USA), aGUI-based software was developed for online imager con-trol, CBF view, CBF analysis in selected ROIs, and dataexporting as shown in Fig. 4(a). The imager control is toconfigure the CCD camera parameters (e.g., exposure time,frame rate, etc.), focus the lens, and switch the light source.The CBF images can also be exported for offline processing.For different applications, three laser speckle contrast anal-ysis algorithms were implemented, i.e., spatial laser specklecontrast analysis (sLASCA),30 temporal laser speckle con-trast analysis (tLASCA),31 and random process estimator

FIG. 4. Demonstration of GUI software and real-time registration of CBF imaging. (a) The graphic interface of the software; (b)–(d) real-time online registrationof raw data. (b) Two ROIs (indicated by the white arrows) selected at the first frame of CBF. (c) The frame without registration shows an obvious shift of thevessel out of the ROIs. (d) The vessel shift was corrected after online image registration.

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

202.120.62.150 On: Sat, 20 Sep 2014 01:50:49

094301-5 Li et al. Rev. Sci. Instrum. 85, 094301 (2014)

FIG. 5. Screenshots of the accompanying video. Real-time registration of raw laser speckle data. (a) The provided video displays simultaneously the changeof CBF map with and without online registration. The end of the video shows the vessels drifting out of the ROIs (b) and the drift was corrected by onlineregistration (c). (Multimedia view) [URL: http://dx.doi.org/10.1063/1.4895657.1].

(RPE) method.32 The software enables preview of the relativeCBF changes over baseline in multiple ROIs as illustrated inFig. 4(a), so that the technician could get real time feedbackduring the operation.

B. Demo of intraoperative CBF monitoringin rat MCAO model

We demonstrate the system in monitoring CBF changesof a rat, during surgery of intraluminal MCAO experiment.The experimental protocols were approved by the AnimalCare and Use Committee of Med-X Research Institute ofShanghai Jiao Tong University.

1. Animal preparation

An adult male Sprague-Dawley rat (Shanghai Slac Lab-oratory Animal Co., Ltd., Shanghai, China), weighing 400 g,was anesthetized with an intraperitoneal injection of 7% chlo-ral hydrate (5 ml/kg). Then the rat was immobilized in thecustomized stereotaxic frame fixed on the OP and the bodytemperature was maintained at 37.0 ± 0.5 ◦C with a heatingpad (FHC, Bowdoinham, USA). The rat skull was exposedby a midline scalp incision and the tissue over the bones wascleaned by a scalpel. A 14 mm × 14 mm cranial window(overlying the cortex) was thinned with a high speed dentaldrill (Strong 90 Micro Motor, Saeshin Precision, Korea) withØ 1.4 mm steel burr until the cortical vessels were clearlyvisible. Saline was applied to prevent the calorification of theskull during the thinning process. After the CBF imaging win-dow was prepared, the rat was then supinely fixed for surgeryof intraluminal MCAO experiment.

2. Permanent MCAO (pMCAO)

The MCAO procedures followed previously publishedprotocols.26 The right common carotid artery (CCA) and ex-ternal carotid artery (ECA) were exposed through a midlineneck incision and the soft tissues were pull apart. The rightCCA and ECA were ligated with 5-0 silk suture lines. Theinternal carotid artery (ICA) was isolated and occluded by acurved microvascular clip temporally. A nylon suture (length:4.0 cm; diameter: 0.36 mm) with 5-mm silicone coating (Bei-jing Sunbio Biotech Co., Ltd, Beijing, China) was advancedinto ICA through a small cut in the right CCA before the bi-furcation between ECA and ICA. A knot was made to fix thefilament and prevent errhysis. After the clip was released, thesuture was advanced up to 18 mm from the bifurcation pointof CCA and then drawn back. We purposely moved the su-ture back and forth. The changes in CBF were imaged and areillustrated in Fig. 6(b).

3. Data analysis and acquisition

We selected the ROIs including two major arteries fromboth contralateral and ipsilateral cortices. Baseline CBF wasassessed with the averaged contrast values at 1-min tempo-ral interval before the surgery. The real time relative CBFchanges were shown in a separate window, so that the tech-nician could observe the CBF responses during the surgery.

III. RESULTS

Besides the standard surgery of intraluminal MCAOexperiment, we purposely manipulated the suture inser-tion depth to test the CBF responses. Figs. 6(b) and

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

202.120.62.150 On: Sat, 20 Sep 2014 01:50:49

094301-6 Li et al. Rev. Sci. Instrum. 85, 094301 (2014)

FIG. 6. Demonstration of the rat model of MCAO. (a) 2D CBF and the ROIs; (b) dynamic CBF changes in the two selected ROIs during MCAO modelexperiment. CBF values are presented with average of measurements during 20-s periods every 2 min as percent of baseline before surgery. (c) The keyoperations are designated, i.e., (I) ligating the right CCA and ECA, (II) clipping the right ICA, (III) inserting the suture, (IV) drawing back the suture, (V)inserting the suture, (VI) drawing back the suture, (VII) inserting the suture, and (VIII) drawing back the suture.

7 (Multimedia view) showed the dynamic changes of averageCBF in two selected ROIs (ROI0, contralateral; ROI1, ipsilat-eral) throughout the experiment.

Ligating the right CCA and ECA resulted in a decreaseof CBF in ROI0 to ∼75% which recovered to the baseline be-

fore dropping to ∼66% after clipping the right ICA. The CBFin ROI1 dropped less but recovered faster than that in ROI0during the same period. Right after the suture was insertedinto the right CCA, there was a sharp drop of CBF in ROI0.It is shown that the CBF had an immediate reperfusion in

FIG. 7. A screenshot of the accompanying video. Dynamic changes of CBF during surgery of intraluminal MCAO experiment. For comparison, the videos andthe animation were included into the same window by post-processing. (Multimedia view) [URL: http://dx.doi.org/10.1063/1.4895657.2].

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

202.120.62.150 On: Sat, 20 Sep 2014 01:50:49

094301-7 Li et al. Rev. Sci. Instrum. 85, 094301 (2014)

contralateral area after the suture was drawn back from theright CCA, which notably recovered to more than 110% ofbaseline. We further advanced and then drew back the fila-ment along the CCA, and the CBF imaging could respondto the operation accurately. The recovery of CBF in ROI0 be-came slower in subsequent filament withdrawals, which mightbe due to the contralateral injury from ischemia. In contrast,ipsilateral CBF remained stable at the baseline level after clip-ping the right ICA. These results suggest that this system iscapable of monitoring the CBF intraoperatively.

IV. CONCLUSIONS

In this paper, we presented a design of system for intra-operative real-time monitoring of CBF in rodent experiments.The system included a specially designed operation bench, anLSCI imager, and CBF imaging software. The imager couldbe installed either under or above the bench so as to mon-itor the CBF during the surgery with the animals in eithersupine or prone position. CBF acquisition, analysis, preview,and data output at selected ROIs could be easily controlledwith a GUI software. Real-time registration of raw speckleimages was implemented to correct the motion artifacts.

Though we demonstrated the intraoperative CBF moni-toring in rodent model of MCAO, our imaging system couldalso be readily applied to other experimental studies, such asthe induction of subarachnoid hemorrhage (SAH)33 for hem-orrhagic stroke study, and animal model vascular conduits.34

And clinically, it could also be translated to potential ap-plications including carotid endarterectomy,5 extracranial-intracranial bypass grafts,7 or revascularization,8 etc.

ACKNOWLEDGMENTS

This work is partly supported by National Natural Sci-ence Foundation of China (No. 61371018).

1J. Koizumi, Y. Yoshida, T. Nakazawa, and G. Ooneda, Jpn. J. Stroke 8, 1–8(1986).

2Y. Li, S. Zhu, L. Yuan, H. Lu, H. Li, and S. Tong, J. Biomed. Opt. 18,076024–076024 (2013).

3E. Farkas, P. G. Luiten, and F. Bari, Brain Res. Rev. 54, 162–180 (2007).4H. Wang, X. Yang, Z. Wang, Z. Deng, H. Gong, and Q. Luo, J. Biomed.Opt. 17, 0612071–0612079 (2012).

5G. Grubhofer, W. Plöchl, M. Skolka, M. Czerny, M. Ehrlich, and A. Lass-nigg, Anesth. Analg. 91, 1339–1344 (2000).

6E. Kapsalaki, G. Lee, J. Robinson III, A. Grigorian, and K. Fountas, J. Clin.Neurosci. 15, 153–157 (2008).

7J. Woitzik, P. Horn, P. Vajkoczy, and P. Schmiedek, J. Neurosurg. 102, 692–698 (2005).

8N. Hecht, J. Woitzik, J. P. Dreier, and P. Vajkoczy, Neurosurg. Focus 27,E11 (2009).

9F. Calamante, D. L. Thomas, G. S. Pell, J. Wiersma, and R. Turner, J. Cereb.Blood Flow Metab. 19, 701–735 (1999).

10W.-D. Heiss, R. Graf, K. Wienhard, J. Löttgen, R. Saito, T. Fujita, G. Ros-ner, and R. Wagner, J. Cereb. Blood Flow Metab. 14, 892–902 (1994).

11M. Wintermark, M. Reichhart, J. P. Thiran, P. Maeder, M. Chalaron, P.Schnyder, J. Bogousslavsky, and R. Meuli, Ann. Neurol. 51, 417–432(2002).

12E. Hungerhuber, S. Zausinger, T. Westermaier, N. Plesnila, and R. Schmid-Elsaesser, J. Neurosci. Methods 154, 109–115 (2006).

13U. Dirnagl, B. Kaplan, M. Jacewicz, and W. Pulsinelli, J. Cereb. BloodFlow Metab. 9, 589–596 (1989).

14N. H. Boyle, P. C. Roberts, B. Ng, H. Berkenstadt, A. McLuckie, R. J.Beale, and R. C. Mason, Crit. Care 3, 95 (1999).

15V. J. Srinivasan, S. Sakadžic, I. Gorczynska, S. Ruvinskaya, W. Wu, J. G.Fujimoto, and D. A. Boas, Opt. Express 18, 2477–2494 (2010).

16M. Xu and L. V. Wang, Rev. Sci. Instrum. 77, 041101 (2006).17A. K. Dunn, H. Bolay, M. A. Moskowitz, and D. A. Boas, J. Cereb. Blood

Flow Metab. 21, 195–201 (2001).18H. Cheng and T. Q. Duong, Opt. Lett. 32, 2188–2190 (2007).19H. Cheng, Q. Luo, Z. Wang, H. Gong, S. Chen, W. Liang, and S. Zeng,

Appl. Opt. 42, 5759–5764 (2003).20B. Kruijt, H. De Bruijn, H. Sterenborg, and D. Robinson, Lasers Med. Sci.

21, 208–212 (2006).21R. Bray, K. Forrester, J. Reed, C. Leonard, and J. Tulip, J. Orthop. Res. 24,

1650–1659 (2006).22Y. C. Huang, T. L. Ringold, J. S. Nelson, and B. Choi, Lasers Surg. Med.

40, 167–173 (2008).23N. Hecht, J. Woitzik, S. König, P. Horn, and P. Vajkoczy, J. Cereb. Blood

Flow Metab. 33, 1000–1007 (2013).24A. K. Dunn, A. Devor, H. Bolay, M. L. Andermann, M. A. Moskowitz, A.

M. Dale, and D. A. Boas, Opt. Lett. 28, 28–30 (2003).25M. Li, P. Miao, Y. Zhu, and S. Tong, J. Biomed. Opt. 16, 086011–086011

(2011).26M. Sasaki, O. Honmou, and J. D. Kocsis, Neural Cell Transplantation

(Springer, 2009), pp. 187–195.27O. Engel, S. Kolodziej, U. Dirnagl, and V. Prinz, J. Vis. Exp. 47, 2423

(2011).28F. Arnberg, J. Lundberg, M. Söderman, P. Damberg, and S. Holmin, Stroke

43, 2437–2443 (2012).29P. Miao, A. Rege, N. Li, N. V. Thakor, and S. Tong, IEEE Trans. Biomed.

Eng. 57, 1152–1157 (2010).30J. D. Briers and S. Webster, J. Biomed. Opt. 1, 174–179 (1996).31H. Cheng, Q. Luo, S. Zeng, S. Chen, J. Cen, and H. Gong, J. Biomed. Opt.

8, 559–564 (2003).32P. Miao, N. Li, N. V. Thakor, and S. Tong, Opt. Express 18, 218–236

(2010).33S. Feiler, B. Friedrich, K. Schöller, S. C. Thal, and N. Plesnila, J. Neurosci.

Methods 190, 164–170 (2010).34M. J. Byrom, P. G. Bannon, G. H. White, and M. K. Ng, J. Vasc. Surg. 52,

176–195 (2010).

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP:

202.120.62.150 On: Sat, 20 Sep 2014 01:50:49