Abstract

Background: Intraoperative monitoring plays a crucial role in reducing complications during neurosurgical procedures. However, effective methods to detect brain tissue viability changes due to blood flow alterations remain unsolved. Electrophysiological techniques, such as motor evoked potentials (MEPs), and fluorescent angiography using indocyanine green, are the primary methods for intraoperative assessment. Real-time intraoperative monitoring is essential for ensuring safe neurosurgical interventions. This study aims to develop a non-contact imaging system for brain tissue surface tissue oxygen saturation (StO2) using red-green-blue (RGB) imaging based on diffuse reflectance spectroscopy.

Methods: Twelve patients with cerebrovascular diseases who underwent craniotomy were included. Six patients had Moyamoya disease, while the remaining six had unruptured cerebral aneurysms. StO2 was monitored in all patients using an RGB camera during surgery.

Results: In Moyamoya disease cases, superficial temporal artery (STA)-middle cerebral artery bypass and encephalo-myo-synangiosis were performed. A significant increase in StO2 was observed after STA release, correlating with cerebral hyperperfusion syndrome as evaluated by ¹⁵O-Positron Emission Tomography scans 1 day post-surgery. In cerebral aneurysm cases, StO2 alterations were noted during internal carotid artery temporary occlusion, potentially impacting MEP outcomes. The effects of various intraoperative parameters on StO2 were evaluated.

Conclusion: Real-time monitoring of StO2 using a highly versatile RGB camera mounted on the side scope of any surgical microscope, regardless of model, is a promising approach for enhancing the safety and efficacy of neurosurgical interventions. By capturing real-time changes in tissue oxygenation, this method may aid in predicting postoperative complications and preventing ischemic events.

Keywords: Brain tissue oxygen saturation, Cerebral hyperperfusion syndrome, Diffuse reflectance spectroscopy, Intraoperative real-time monitoring, Red-green-blue camera

INTRODUCTION

There is currently no established method for neurosurgeons to monitor changes in brain tissue viability caused by blood flow alterations during parent vessel occlusion in complex cerebral aneurysms or hyperperfusion states following bypass surgery in Moyamoya disease. The ability to monitor these changes in real time is crucial for preventing complications during surgery. However, this remains a significant challenge. At present, electrophysiological techniques such as motor-evoked potentials (MEPs)[ 1 , 12 ] and fluorescent angiography using indocyanine green (ICG)[ 14 ] are the primary methods used for intraoperative assessment. While laser speckle imaging, which visualizes hemoglobin movement within capillaries, has been considered a potential alternative, it requires laser irradiation, has a limited range, and cannot assess brain tissue viability.[ 10 ] Similarly, there is no established real-time monitoring technique for detecting cerebral hyperperfusion syndrome (CHS) during surgery.

To address this issue, diffuse reflectance spectroscopic analysis has been proposed as a method for monitoring intraoperative brain surface tissue oxygen saturation (StO₂).[ 15 ] This technique leverages the optical properties used as intrinsic optical signals of cerebral tissues, which are influenced by cerebral blood flow (CBF), blood volume, and hemoglobin StO₂.

High-spectrum cameras originally developed for National Aeronautics and Space Administration’s (NASAs) Mars exploration have been utilized in revascularization and epilepsy surgeries to analyze these optical signals.[ 5 , 11 ] However, their high cost, specialized equipment requirements, and incompatibility with surgical microscopes limit their practical use. Furthermore, these methods rely on postoperative analysis rather than real-time monitoring.

In our previous animal studies, we demonstrated that an inexpensive and highly versatile red-green-blue (RGB) camera could detect brain surface StO₂ using the same principles.[ 7 ] This suggests the potential of this method for assessing pathophysiological conditions and brain tissue viability loss. In this study, we integrated an inexpensive and highly versatile RGB camera into an existing surgical microscope and extracted the projected image to generate a color-coded visualization of the brain surface StO₂ during surgery in real time. This study aimed to develop a real time monitoring system to detect alterations in StO₂ as a brain tissue viability due to blood flow changes during surgery.

MATERIALS AND METHODS

The subjects were 12 patients (male, n = 3; female, n = 9; mean age 55.7 ± 18.2 years; age range 13–77 years) with cerebrovascular disease who underwent craniotomy under general anesthesia at the Department of Neurosurgery, Yamagata University Hospital from January 2020 to January 2023. Six patients had moyamoya disease and six had unruptured cerebral aneurysms. All patients underwent real-time monitoring of StO₂ acquired using an RGB camera during surgery.

For patients with moyamoya disease, we performed superficial temporal artery (STA)-middle cerebral artery (MCA) bypass and encephalo-myo-synangiosis (EMS) with dural pedicle insertion. ¹⁵O-Positron Emission Tomography (PET) was performed a day after surgery, and CBF images were analyzed using a 3D stereotaxic region of interest (ROI) template (3DSRT).[ 13 ] We evaluated the contralateral ratio at the site where the increase in blood flow was most significant compared with the contralateral side. CHS was defined as local hyperperfusion around the anastomosis site where the CBF contralateral ratio is increased on ¹⁵O-PET, and symptoms of focal neurological deficits, strong headaches, or convulsions.[ 2 ] For cases of cerebral aneurysms, alterations in StO₂ and MEP were evaluated when the internal cerebral artery (ICA) was temporary occluded in the neck or intracranially.

The following intraoperative parameters that may affect StO₂ were controlled with the cooperation of the anesthesiologist: percutaneous oxygen (SpO₂), pressure of oxygen in arterial blood (PaO₂), pressure of carbon dioxide in arterial blood (PaCO₂), systolic blood pressure (SBP), and hemoglobin.

Real-time monitoring of brain tissue StO2 using an RGB camera

We used an inexpensive and versatile RGB charge coupled device (CCD) color camera (DFK-21BU618, IMAGING SOURCE, Germany) and a neurosurgical microscope (M525 OH4, Leica, Germany), with a distance of 280 mm from the brain surface during monitoring and ×2.8 zoom. The imaging settings are illustrated in Figure 1a . This RGB camera can be attached to the side scope of any surgical microscope, regardless of model [ Figure 1b ]. The principle of measurement is the method of estimating StO₂ (%) from light reflectance images based on what we have reported.[ 7 ] The amount of oxygenated and deoxygenated hemoglobin was analyzed by the color values of RGB color image images [ Figure 1c ] using a non-contact microscope based on the theory of light propagation within biological tissues. The oxygenated hemoglobin concentration (CHbO) can be determined by analyzing the three bands of red, green, and blue acquired RGB images in real time using an image-processing program created using MATLAB® (MathWorks, USA). The image was converted to a deoxygenated hemoglobin concentration (CHbR) image. The CHbO and CHbR values were obtained for each pixel. These estimation formulas are established to simulate light propagation within brain tissue and diffuse reflectance spectra obtained from the brain surface under different combinations of CHbO and CHbR based on multiple regression analyses of numerous statistically determined data sets. The total hemoglobin concentration (CHbT) was obtained by adding the CHbO and CHbR values, and the brain surface StO₂ was calculated from the relationship StO₂ (%) = 100 × (CHbO/CHbT).

Figure 1:

Real-time brain tissue oxygen saturation monitoring system using a versatile red-green-blue (RGB) camera. (a) Representative photo of real-time brain tissue oxygen saturation monitoring system during surgery. (b) Photo of a versatile RGB camera (white dot square). (c) Representative RGB digital color image. (d) Representative brain tissue oxygen saturation image. The white square indicates ROI for the analysis of brain surface StO2.

We set the ROI, which could be freely moved during surgery and observed the changes in StO₂ as a trend ([ Figure 1d ] White square).

For cases of moyamoya disease, we evaluated the rate of change in StO₂ as ΔStO₂ in the ROI near the site of anastomosis, before and after STA-MCA bypass. For cerebral aneurysms, an ROI was set in the frontal lobe, and we observed alterations in StO₂ before and during temporary ICA occlusion. Furthermore, we evaluated the rate of change in StO₂ as ΔStO₂ in the ROI before and during the temporary ICA occlusion.

15O-PET

The Siemens Biograph micro-computed tomography (CT) apparatus (software version Syngo VG60A; Siemens Healthcare, Erlangen, Germany) consisted of a PET detector with four rings, 48 detector blocks in each ring, and lutetium oxyorthosilicate (LSO) crystals of 4 × 4 × 20 mm in a 13 × 13 array coupled to a 2 × 2 photomultiplier tube (PMT) array in each detector block. This provides an axial PET field of view (FOV) of 22.1 cm. The transaxial FOV was 70 cm. The diameter of the detector ring was 84.2 cm. The time coincidence window was 4.1 ns, and the energy window was 435–650 keV. Integrated 64-slice CT was used for attenuation correction of the PET data. CBF was determined while the participant continuously inhaled C15O₂ through a mask and was calculated using the steady-state method.[ 9 ]

Image analyses

A total of 318 constant ROIs were automatically placed in both the cerebral and cerebellar hemispheres using 3DSRT with SPM2 (Fujifilm RI Pharma Co., Ltd., Tokyo, Japan).[ 13 ] The 3DSRT is a fully automated regional CBF quantification software program developed in the Montreal Neurological Institute (MNI) space. The ROIs were grouped into 10 segments (callosomarginal, pericallosal, precentral, central, parietal, angular, temporal, posterior, hippocampus, and cerebellar) in each hemisphere, according to the arterial supply. The affected-to-unaffected-side (A/U) ratio was calculated for each segment. The lowest (preoperative) and highest (postoperative) ratio among the segments is shown in Table 1 .

Table 1:

Summary of patients with real-time monitoring of brain tissue oxygen saturation by RGB camera adopted microscope

The present study was approved by the Ethics Committee of Faculty of Medicine Yamagata University based on ethical guidelines. Informed consent was obtained from all the participants.

RESULTS

The clinical characteristics and intraoperative parameters are shown in Table 1 . We maintained SpO₂, PaO₂, PaCO₂, SBP, and hemoglobin within the appropriate range.

In cases of moyamoya disease, the mean age was 36.5 ± 19.7, and two of six patients were male. CHS was observed in three out of six cases. In the CHS group, the average ΔStO₂ value was >10%. In contrast, in non-CHS, the average ΔStO₂ value was <10%. Although there were no differences in preoperative CBF A/U ratio (¹⁵O-PET) between the CHS and non-CHS groups, the postoperative CBF A/U ratio in the CHS groups demonstrated a tendency to have a higher value than that in non-CHS groups [ Table 1 ].

In cases of cerebral aneurysms, the mean age was 69.8 ± 5.2, and one of the six patients was male. The aneurysm was located in the internal carotid-posterior communicating artery (IC-PC) in five cases and the anterior communicating artery (Acom) in one case. MEP reduction was observed during temporary ICA occlusion in one of the six cases. The ΔStO₂ was −8.7%. The average of the other five cases was −2.1% ± 1.2(1.1−4.0%). In one case in which MEP decline was observed, the patient had a large ΔStO₂ and the decline in ΔStO₂ had not stopped.

Representative cases

Moyamoya disease

CHS (+) Case 1

The patient was a 46-year-old female with bilateral stage III moyamoya disease. She developed a transient ischemic attack (TIA). She underwent a left STA-MCA bypass and EMS on the left side. StO₂ increased from 60% to 77% after STA release, and ΔStO₂ was 28.3% [ Figures 2a - e ]. ¹⁵O-PET CBF on the day after surgery showed an increase in the contralateral ratio around the left frontal lobe adjacent to the bypass anastomosis site [ Figures 2f - h ]. Since the patient had motor aphasia, we diagnosed her with CHS.

Figure 2:

Representative case – Case 1: Moyamoya disease cerebral hyperperfusion syndrome (+), left superficial temporal artery-middle cerebral artery (STA-MCA) bypass + encephalo-myo-synangiosis. (a) Brain tissue oxygen saturation image before STA-MCA bypass. The white square indicates ROI for the surface tissue oxygen saturation (StO₂) analysis. (b) Photo of brain tissue oxygen saturation image after STA-MCA bypass. The white square indicates ROI for StO₂ analysis. An increase in StO₂ was observed. (c) Graph of the trend of StO₂ before and after STA-MCA bypass. An increase of StO₂ was observed just after STA release (black arrow). (d and e): Preoperative (d) and postoperative (e) magnetic resonance angiogram (MRA). Postoperative MRA shows an increased left STA signal (white arrow). (f and g) Preoperative (f) and postoperative (g) ¹⁵O-Positron emission tomography (PET) (cerebral blood flow [CBF]). Postoperative ¹⁵O-PET shows increased CBF in the left frontal lobe. (h) The ROI image by 3D stereotaxic ROI template, which analyzes the A/U ratio, shows that the hyperperfusion area is in the precentral segment (b). ROI: Region of interest.

CHS (−) Case 2

The patient was a 13-year-old female with moyamoya disease (stage III on the right side and stage II on the left side). She developed a TIA. She underwent a right STA-MCA bypass and EMS on the right side. Her StO₂ value increased from 24% tO₂6% after STA release, and her ΔStO₂ value was 8.3% [ Figures 3a - c ]. No CHS was observed during the postoperative course [ Figures 3d and e ].

Figure 3:

Representative case – Case 2: Moyamoya disease cerebral hyperperfusion syndrome (−), right superficial temporal artery-middle cerebral artery (STA-MCA) bypass + encephalomyo-synangiosis. (a) Graph of the trend of surface tissue oxygen saturation (StO₂) before and after STA-MCA bypass. A slight increase of StO₂ was observed just after STA release (black arrow). (b and c) Preoperative (b) and postoperative (c) magnetic resonance angiogram (MRA). Postoperative MRA shows an increased right STA signal (white arrow). (d and e) Preoperative (d) and postoperative (e) ¹⁵O-Positron emission tomography (PET) (cerebral blood flow [CBF]). Postoperative ¹⁵O-PET indicates a slightly increased CBF in the right frontal lobe.

Cerebral aneurysms

MEP decline (+) Case 8

The patient was a 77-year-old woman with a right IC-PC unruptured aneurysm [ Figure 4a ]. A decrease in MEP was observed 1 min after temporary occlusion of the ICA and disappeared within 1.5 min. The StO₂ value did not stop decreasing and gradually decreased until MEP declined [ Figure 4b ]. After release, the MEP recovered and the postoperative course was uneventful [ Figure 4c ]. The posterior communicating artery was poorly developed, and the Acom was unclear.

Figure 4:

Representative case – Case 8: Right internal carotid-posterior communicating (IC-PC) unruptured aneurysm, motor-evoked potential (MEP) decline (+) (a) Preoperative three-dimensional computed tomographic angiography The black arrow indicates the right IC-PC unruptured aneurysm. (b) Graph of the trend of surface tissue oxygen saturation (StO2) before and after the right internal cerebral artery (ICA) temporary occlusion just after the ICA was occluded (black arrow), StO2 gradually decreased and continued to decrease. MEP declined at 60 s after ICA occlusion (arrowhead). Finally, the StO2 increased just after occlusion was released at 80 s (dotted arrow). (c) MEP monitoring. A decline in MEP was observed 1 min after right ICA clamping and disappeared within one and a half min. After declamping of the right ICA at 80 s, MEP recovered completely.

MEP change (−) Case 7

The patient was a 69-year-old woman with a right IC-PC unruptured aneurysm. No decrease in MEP was observed during temporary ICA occlusion, and the decrease in StO₂ during occlusion was transient and recovered, with a maximum change rate of −4% [ Figure 5 ].

Figure 5:

Representative case – Case 7: Right internal carotid-posterior communicating unruptured aneurysm, motor evoked potential decline (−). Graph of the trend of surface tissue oxygen saturation (StO2) before and after the right internal cerebral artery (ICA) temporary occlusion. The decrease in StO2 just after the right ICA temporary occlusion (black arrow) was transient and recovered at 20 s after occlusion (dot arrow), with a maximum change rate of −4%.

DISCUSSION

The system used in this study employs a general-purpose RGB camera to estimate brain surface StO₂ from spectral reflectance images, and its reliability has already been demonstrated in animal experiments using rats.[ 7 ] Furthermore, as demonstrated in the results of this study, the system allows real-time visualization of StO₂ color maps and facilitates trend analysis within specific ROIs, making it applicable for intraoperative real-time monitoring. In addition, since the RGB camera can be mounted on the side scope of any surgical microscope, regardless of model, it holds significant potential for practical use in actual surgical settings.

There have been several reports of methods to observe StO₂ on intraoperative brain surface images using a spectroscopic analysis. Mori et al.[ 5 ] reported the use of a hyperspectral camera (HSC), developed by NASA for Mars exploration, to provide continuous imaging spectroscopic quantification of hemodynamic responses in the brain, with high reliability.

Pichette et al.[ 11 ] found that HSC could capture changes in StO₂ that corresponded to the area of spikes in intraoperative electroencephalography (EEG) in cases of epilepsy. Although these HSCs may have potential uses, they are limited due to their high cost, and because they are applied postoperatively and do not have real-time monitoring capabilities.

The usefulness of the fluorescent dye ICG has been reported as a method for intraoperatively assessing the risk of CHS during revascularization for ischemic cerebrovascular diseases, such as moyamoya disease.[ 14 ] This dye is administered intravenously, and it is possible to observe changes in blood flow for several minutes; however, continuous monitoring during surgery is not possible.

There have been some reports in the past about monitoring based on changes in the brain surface tissue during surgery, as in this study. A method that can measure brain temperature changes using an infrared imaging device in revascularization surgery for moyamoya disease and capture changes in CBF after revascularization[ 8 ] and a method of observing metabolism from intraoperative brain surface images using a spectroscopic analysis method using an HSC[ 4 ] have been reported to be useful for predicting postoperative CHS. However, these systems are costly, have low versatility, and are complicated to set up. Furthermore, the observation range is limited, the analysis results cannot be grasped in real time during surgery, and they are based on a postoperative analysis; therefore, they cannot yet be used for intraoperative monitoring, and further development is needed.

A previous study monitoring cerebral StO₂ transcranially using near-infrared spectroscopy (NIRS) during carotid endarterectomy reported that an increase in StO₂ after reperfusion correlated with CHS.[ 3 ] This suggests that StO₂ is a useful indicator of cerebral circulation dynamics. In our study, all cases of CHS after revascularization for moyamoya disease showed a ΔStO₂ increase of more than 10% after STA release. Thus, the degree of ΔStO₂ increase following revascularization may serve as a predictor of CHS. Furthermore, ischemic complications occurred in cases where StO₂ decreased during internal carotid artery blockage and did not stop decreasing.[ 6 ] In our study, we also found a decrease in MEP in cases in which ΔStO₂ did not stop decreasing during parent artery occlusion, suggesting that earlier changes may have been detected.

On the other hand, in cerebral aneurysm surgery, it is necessary to prevent brain tissue damage due to ischemia when parent arteries are occluded or temporarily blocked, but this is currently judged by electrophysiological monitoring using the MEP.[ 1 , 12 ] These monitoring methods are fixed-point observations rather than continuous measurements, so they may suddenly decrease or disappear at some point. This can lead to a delayed response; however, if viability can be evaluated in real-time based on oxygen saturation of a wide range of brain surface tissue during surgery, as in this study, it will be possible to capture changes at an earlier stage, which may further reduce ischemic complications.

The limitation of this study is the small number of cases, which limits the statistical robustness and generalizability of the findings. Moreover, the system used in this study relies on relative changes in StO₂, as absolute values are not yet reliably validated, reducing its standalone diagnostic utility. We also need to be aware of the system, which only captures surface-level oxygenation and may not reflect deeper tissue perfusion. Finally, there is no direct intraoperative comparison with established modalities such as NIRS or laser speckle imaging, which would have strengthened the validation of this technique.

In conclusion, in this study, by simply attaching a highly versatile RGB camera to an existing surgical microscope, it was possible to monitor the brain surface StO₂, which is thought to correlate with the behavior of cerebral circulation, in real time during surgery. By capturing changes in blood flow during surgery as changes in StO₂, it might be possible to predict CHS after revascularization and prevent ischemic complications due to parent artery blockage/occlusion earlier. It is believed that this could be a useful form of intraoperative monitoring to safely perform neurosurgery.

CONCLUSION

This study demonstrated that intraoperative real-time brain tissue oxygen saturation imaging using the method we developed to estimate brain surface StO₂ from spectral reflectance images using a general-purpose RGB camera may be useful as an intraoperative monitoring system for evaluating the viability of brain tissue during temporary occlusion of major cerebral arteries during surgery and for predicting CHS after revascularization. We would like to make additional enhancements to increase its reliability.

Ethical approval:

The research/study approved by the Institutional Review Board at Yamagata University Faculty of Medicine, number 2022-16, dated May 13, 2022.

Declaration of patient consent:

The authors certify that they have obtained all appropriate patient consent.

Financial support and sponsorship:

Japan Society for the Promotion of Science Grant-in-Aid for Scientific Research KAKENHI (Grant Number 20H04513).

Conflicts of interest:

There are no conflicts of interest.

Use of artificial intelligence (AI)-assisted technology for manuscript preparation:

The authors confirm that there was no use of artificial intelligence (AI)-assisted technology for assisting in the writing or editing of the manuscript and no images were manipulated using AI.

Disclaimer

The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of the Journal or its management. The information contained in this article should not be considered to be medical advice; patients should consult their own physicians for advice as to their specific medical needs.

References

1. Cedzich C, Taniguchi M, Schafer S, Schramm J. Somatosensory evoked potential phase reversal and direct motor cortex stimulation during surgery in and around the central region. Neurosurgery. 1996. 38: 962-70

2. Fujimura M, Kaneta T, Mugikura S, Shimizu H, Tominaga T. Temporary neurologic deterioration due to cerebral hyperperfusion after superficial temporal artery-middle cerebral artery anastomosis in patients with adult-onset moyamoya disease. Surg Neurol. 2007. 67: 273-82

3. Iwaki K, Takagishi S, Arimura K, Murata M, Chiba T, Nishimura A. A novel hyperspectral imaging system for intraoperative prediction of cerebral hyperperfusion syndrome after superficial temporal artery-middle cerebral artery anastomosis in patients with moyamoya disease. Cerebrovasc Dis. 2021. 50: 208-15

4. Kamenskaya OV, Loginova IY, Lomivorotov VV. Brain oxygen supply parameters in the risk assessment of cerebral complications during carotid endarterectomy. J Cardiothorac Vasc Anesth. 2017. 31: 944-9

5. Mori M, Chiba T, Nakamizo A, Kumashiro R, Murata M, Akahoshi T. Intraoperative visualization of cerebral oxygenation using hyperspectral image data: A two-dimensional mapping method. Int J Comput Assist Radiol Surg. 2014. 9: 1059-72

6. Moritz S, Kasprzak P, Arlt M, Taeger K, Metz C. Accuracy of cerebral monitoring in detecting cerebral ischemia during carotid endarterectomy: A comparison of transcranial Doppler sonography, near-infrared spectroscopy, stump pressure, and somatosensory evoked potentials. Anesthesiology. 2007. 107: 563-9

7. Mustari A, Kanie T, Kawauchi S, Sato S, Sato M, Kokubo Y. In vivo evaluation of cerebral hemodynamics and tissue morphology in rats during changing fraction of inspired oxygen based on spectrocolorimetric imaging technique. Int J Mol Sci. 2018. 19: 491

8. Nakagawa A, Fujimura M, Arafune T, Sakuma I, Tominaga T. Clinical implications of intraoperative infrared brain surface monitoring during superficial temporal artery-middle cerebral artery anastomosis in patients with moyamoya disease. J Neurosurg. 2009. 111: 1158-64

9. Okazawa H, Yamauchi H, Sugimoto K, Takahashi M, Toyoda H, Kishibe Y. Quantitative comparison of the bolus and steady-state methods for measurement of cerebral perfusion andoxygen metabolism: Positron emission tomography study using 15O-gas and water. J Cereb Blood Flow Metab. 2001. 21: 793-803

10. Paul JS, Luft AR, Yew E, Sheu FS. Imaging the development of an ischemic core following photochemically induced cortical infarction in rats using laser speckle contrast analysis (LASCA). Neuroimage. 2006. 29: 38-45

11. Pichette J, Laurence A, Angulo L, Lesage F, Bouthillier A, Nguyen DK. Intraoperative video-rate hemodynamic response assessment in human cortex using snapshot hyperspectral optical imaging. Neurophotonics. 2016. 3: 045003

12. Sasaki T, Suzuki K, Sakuma J, Sato T, Kasuya H, Ando H. Prevention of ischemic complications after aneurysm surgery using various monitoring method. Surg Cereb Stroke (JPN). 2009. 37: 79-86

13. Takeuchi R, Matsuda H, Yoshioka K, Yonekura Y. Cerebral blood flow SPET in transient global amnesia with automated ROI analysis by 3DSRT. Eur J Nucl Med Mol Imaging. 2004. 31: 578-89

14. Uda K, Araki Y, Muraoka S, Ota S, Wada K, Yokoyama K. Intraoperative evaluation of local cerebral hemodynamic change by indocyanine green videoangiography: Prediction of incidence and duration of postoperative transient neurological events in patients with moyamoya disease. J Neurosurg. 2018. 130: 1367-75

15. Yoshida K, Nishidate I, Ishizuka T, Kawauchi S, Sato S, Sato M. Multispectral imaging of absorption and scattering properties of in vivo exposed rat brain using a digital red-green-blue camera. J Biomed Opt. 2015. 20: 051026