Sage Journals: Discover world-class research

Abstract

Currently, there is no adequate technique for intraoperative monitoring of cerebral blood flow (CBF). To evaluate laser speckle imaging (LSI) for assessment of relative CBF, LSI was performed in 30 patients who underwent direct surgical revascularization for treatment of arteriosclerotic cerebrovascular disease (ACVD), Moyamoya disease (MMD), or giant aneurysms, and in 8 control patients who underwent intracranial surgery for reasons other than hemodynamic compromise. The applicability and sensitivity of LSI was investigated through baseline perfusion and CO₂ reactivity testing. The dynamics of LSI were assessed during bypass test occlusion and flow initiation procedures. Laser speckle imaging permitted robust (pseudo-) quantitative assessment of relative microcirculatory flow and standard bypass grafting resulted in significantly higher postoperative baseline perfusion values in ACVD and MMD. The applicability and sensitivity of LSI was shown by a significantly reduced CO₂ reactivity in ACVD (9.6 ± 9%) and MMD (8.5 ± 8%) compared with control (31.2 ± 5%; P < 0.0001). In high- and intermediate-flow bypass patients, LSI was characterized by a dynamic real-time response to acute perfusion changes and ultimately confirmed a sufficient flow substitution through the bypass graft. Thus, LSI can be used for sensitive and continuous, non-invasive real-time visualization and measurement of relative cortical CBF in excellent spatial-temporal resolution.

Keywords

cerebrovascular disease EC-IC bypass surgery intraoperative monitoring of cerebral blood flow laser speckle imaging stroke

Introduction

Current intraoperative monitoring techniques lack adequate surveillance of cerebral blood flow (CBF). Monitoring of the cerebral microcirculation is crucial, because tissue viability and clinical outcome depend on restoration of blood flow to an ischemic region in a timely manner.^1,2 At present, intraoperative visualization of vessel perfusion is routinely accomplished by indocyanine green (ICG) video angiography³ and off-line transit-time calculations of the fluorescent ICG dye within the arterial, parenchymal and venous compartments have served as surrogate markers of CBF.^4,5 Recently, a new software tool for color-coded mapping of this transit-time pattern became available and it was suggested that the methodology could be useful for intraoperative monitoring of regional CBF.⁶ While the method is feasible and may be suitable for characterization of perfusion alterations before and after a surgical intervention, it is not an established means for CBF measurement.

The two main techniques for true non-invasive assessment of relative cortical CBF (= relative cortical perfusion) are laser doppler flowmetry (LDF) and laser doppler imaging (LDI). However, both methods lack the combination of a high spatial and high temporal resolution. While LDF is limited by registering the blood flow only at a single point, LDI does not visualize the blood flow in real time because of the time required for surface scanning and data processing.⁷

Laser speckle imaging (LSI) is an experimentally validated, true full-field optical imaging technique that avoids the need for surface scanning. Light from a laser is diverged by a lens to illuminate the area under investigation and a charge-coupled device (CCD) camera images the illuminated area for real-time generation of a two-dimensional color-coded map of blood flow.^7-9 Preliminary investigations by others and our group have shown the general feasibility of intraoperative LSI during neurosurgical procedures.^10,11 However, both studies were limited by a small sample size (n= 3) and lacking evidence of routine clinical applicability.

The purpose of this study was to evaluate intraoperative LSI as a tool for dynamic real-time assessment of relative cortical CBF in the human brain by determining normal values of LSI-specific perfusion parameters and the change of these parameters in cases of known perfusion anomalies.

MATERIALS AND METHODS

Study Design

The study (EA2/139/12) was approved by the local ethics committee of the Charité University Hospital in Berlin, Germany, and included 38 patients who underwent 61 LSI procedures in our institution between March 2010 and April 2011. Patient data are presented in Table 1. Intraoperative LSI was performed following a fixed protocol in all patients. Twenty-three patients received superficial temporal artery to middle cerebral artery (STA-MCA) bypass grafting for treatment of single-photon emission computed tomography (SPECT)-confirmed hemodynamic compromise due to arteriosclerotic cerebrovascular disease (ACVD; n = 9 male, n = 5 female; median age 64 (range 48 to 84)) or Moyamoya disease (MMD; n = 1 male, n = 8 female; median age 35 (range 17 to 59)). Seven patients underwent high-flow (saphenous vein) or intermediate-flow (radial artery) bypass revascularization before trapping of a giant aneurysm in the anterior circulation (n = 2 male, n = 5 female; median age 48 (range 14 to 71)). Eight patients served as hemodynamically healthy controls (n = 7 male, n = 1 female; median age 47 (range 19 to 76)).

Table 1.

Demographic and clinical patient data.

No.	Age, sex	Diagnosis	Surgical procedure.	Localization of LSI imaging field	Test during intraoperative LSI	ICG bypass patency
01	55, m	ACVD	Left STA-MCA	Left perisylvian	Baseline / CO₂	Yes
02	72, m	ACVD	Left STA-MCA	Left perisylvian	Baseline / CO₂	Yes
03	54, f	Giant-ICA Aneurysm	Right IF ECA-M2	Right perisylvian	Test occlusion	Yes
04	48, f	ACVD	Right STA-MCA	Right perisylvian	Baseline / CO₂	Yes
05	57, m	ACVD	Right STA-MCA	Right perisylvian	Baseline / CO₂	Yes
06	26, f	MMD	Left STA-MCA	Left perisylvian	Baseline / CO₂	Yes
07	30, m	MMD	Left STA-MCA	Left perisylvian	Baseline / CO₂	Yes
08	42, m	Giant-AcomA Aneurysm	Right IF STA-A3 db	Right perisylvian	Test occlusion	Yes
09	71, f	Giant-ICA Aneurysm	Right IF ECA-M2	Right perisylvian	Test occlusion	Yes
10	48, f	Giant-ICA Aneurysm	Left HF ECA-M2	Left perisylvian	Test occlusion	Yes
11	68, f	ACVD	Left STA-MCA	Left perisylvian	Baseline / CO₂	Yes
12	55, f	Giant-ICA Aneurysm	Left HF ECA-M2	Left perisylvian	Test occlusion	Yes
13	31, f	MMD	Left STA-MCA + EDS	Left perisylvian	Baseline / CO₂	Yes
14	85, m	ACVD	Right STA-MCA	Right perisylvian	Baseline / CO₂	Yes
15	55, f	MMD	Right STA-MCA	Right perisylvian	Baseline / CO₂	Yes
16	76, f	ACVD	Right STA-MCA	Right perisylvian	Baseline / CO₂	Yes
17	67, m	ACVD	Left STA-MCA	Left perisylvian	Baseline / CO₂	Yes
18	67, f	ACVD	Left STA-MCA	Left perisylvian	Baseline / CO₂	Yes
19	35, f	MMD	Left STA-MCA + EDS	Left perisylvian	Baseline / CO₂	Yes
20	48, m	Giant-AcomA Aneurysm	Right IF MCA-A2 db	Right perisylvian	Test occlusion	Yes
21	63, f	ACVD	Right STA-MCA	Right perisylvian	Baseline / CO₂	Yes
22	37, f	MMD	Right STA-MCA + EMS	Right perisylvian	Baseline / CO₂	Yes
23	53, m	ACVD	Left STA-MCA	Left perisylvian	Baseline / CO₂	Yes
24	65, m	ACVD	Left STA-MCA	Left perisylvian	Baseline / CO₂	Yes
25	17, f	MMD	Right STA-MCA + EMS	Right perisylvian	Baseline / CO₂	Yes
26	14, f	Giant-ICA Aneurysm	Left HF ECA-M2	Left perisylvian	Test occlusion	Yes
27	56, m	ACVD	Right STA-MCA	Right perisylvian	Baseline / CO₂	Yes
28	55, m	ACVD	Left STA-MCA	Left perisylvian	Baseline / CO₂	Yes
29	59, f	MMD	Left STA-MCA + EDS	Left perisylvian	Baseline / CO₂	Yes
30	58, f	MMD	Right STA-MCA + EDS	Right perisylvian	Baseline / CO₂	Yes
C1	19, m	Cavernoma	Excision	Left perisylvian	Baseline / CO₂	N/A
C2	37, f	Epilepsia	Grid implant	Right perisylvian	Baseline / CO₂	N/A
C3	32, m	P3 Aneurysm	P2 Clipping	Right occipital	Baseline / CO₂	N/A
C4	48, m	Glioblastoma	Resection	Left perisylvian	Baseline / CO₂	N/A
C5	65, m	Epilepsia	Temporal lobe resection	Left perisylvian	Baseline / CO₂	N/A
C6	46, m	Epilepsia	Grid implant	Left perisylvian	Baseline / CO₂	N/A
C7	76, m	Lymphoma	Biopsy	Right perisylvian	Baseline / CO₂	N/A
C8	65, m	Glioblastoma	Resection	Left perisylvian	Baseline / CO₂	N/A

Abbreviations: AcomA, anterior communicating artery; ACVD, arteriosclerotic cerebrovascular disease; db, double-barrel bypass; EDS, encephalo-duro-synangiosis; EMS, encephalo-myo-synangiosis; f, female; HF, high-flow; ICA, internal carotid artery; ICG, indocyanine green; IF, intermediate-flow; LSI, laser speckle imaging; m, male; MMD, Moyamoya disease; N/A, not applicable; STA-MCA, superficial temporal artery to middle cerebral artery.

The following LSI-specific test parameters were evaluated:

To determine whether intraoperative LSI following a standardized protocol permits robust assessment of CBF-Flux for (pseudo-) quantitative comparison of relative CBF, we measured cortical baseline perfusion in patients with and without hemodynamic impairment.

The general applicability and sensitivity of LSI was determined by testing the cerebrovascular carbon dioxide (CO₂) reactivity in patients with and without hemodynamic impairment.

The response behavior (dynamics) of LSI for real-time detection of acute perfusion changes was determined during test-occlusion and flow initiation procedures in patients undergoing high- or intermedi ate-flow bypass grafting.

The reliability of assessing bypass graft patency by LSI was evaluated by comparing graft patency assessment by LSI with graft patency assessment by ICG angiography.

Patient Management

All patients were anesthetized with propofol and remifentanil. Heart rate, arterial blood pressure, peripheral oxygen saturation and bladder temperature were registered throughout the surgery. The intraoperative mean arterial pressure (MAP) in patients undergoing bypass surgery for treatment of chronic cerebral hypoperfusion was set between 90 and 100 mm Hg. In patients not suffering from chronic cerebral hypoperfusion, an intraoperative MAP of 80 mm Hg was targeted. During all LSI measurements, MAP was maintained at a constant level and registered simultaneously with the cortical blood flow measurement. In cases where two LSI measurements were performed in one patient during the same surgery, we aimed for an equivalent MAP during each LSI measurement. By default, arterial blood gases were sampled every 30 minutes and the end-expiratory carbon dioxide (CO₂) concentration during surgery was maintained at a level corresponding to an arterial partial pressure of CO₂ (PaCO₂) between 38 and 42 mm Hg. In cases where LSI was used to assess baseline perfusion and CO₂ reactivity in one continuous measurement, additional blood gas samples were obtained during the 2-minute baseline perfusion measurement (PaCO₂ low) and upon completion of the subsequent 5- to 7-minute CO₂ challenge (PaCO₂ high).

All patients received postoperative computerized tomography or magnetic resonance imaging within 24 hours and additionally in cases when new neurologic deficits occurred within the first 10 postoperative days to rule out procedure-related infarction and/or hemorrhage. Bypass patients received CT angiography 24 hours after surgery and digital subtraction angiography on day 7 to assess postoperative bypass patency. Intraoperative bypass patency was assessed by ICG video angiography.¹²

Technical Specifications

The laser speckle imager we used is a 22 × 23 × 8 cm portable class 1 laser device (MoorFLPI, Moor Instruments Ltd., Axminster, UK) with a laser wavelength of 785 nm and power of 50 mW. An integrated CCD camera allows dynamic acquisition of two-dimensional, color-coded maps of perfusion. Physically, LSI measures the same blood flow parameter as conventional LDF in terms of relative blood flow velocity. This blood flow velocity is typically expressed in an arbitrary perfusion unit, commonly termed as Flux (CBF-Flux).

Device Setup

The imager was fixed to a clamped joint and attached to a mobile stand to permit versatile positioning over a desired imaging field. For purposes of data acquisition, the imager was connected to a laptop computer (Dell Computer Cooperation, Round Rock, TX, USA) via standard universal serial bus (USB) and FireWire (IEEE 1394) interfaces (Figure 1). After craniotomy and durotomy, the laser speckle imager was positioned 300 mm above and perpendicular to the exposed cortical area as described previously.¹⁰ Identical focal and zoom settings were maintained within a 5 × 5 cm² imaging field in control and STA-MCA bypass patients and a 7 × 7 cm² imaging field in patients undergoing high- and intermediate-flow bypass revascularization. Direct illumination of the surgical field by light sources other than the laser light was avoided. An adjustable polarizing filter was used to eliminate specular reflection from shiny tissue. Blood clots on the cortical surface were cleared by sterile saline irrigation.

Figure 1.

Device setup for intraoperative laser speckle imaging. Intraoperative setup of the laser speckle device, (1) fixed to a clamped joint (2) and attached to a mobile trolley, which permits free positioning over the surgical field. Standard USB and FireWire interfaces (3) enable a connection to a laptop computer (4) for data recording.

System Configuration

To optimize the measurable perfusion range, sensitivity and signal-to-noise ratio of the LSI measurement, two different exposure times of the CCD camera and two different temporal filters were selected. For measurements of baseline perfusion and CO₂ reactivity in control and STA-MCA bypass patients, relative CBF was recorded at 25 Hz using a temporal filter of 100 frames per image, resulting in a scan rate of 0.25 Hz (768 × 576 pixel) at an exposure time of 8.4 ms for each image. For detection of acute perfusion changes in patients undergoing high- or intermediate-flow bypass grafting, the temporal filter was set to 25 frames per image, resulting in a scan rate of 1 Hz (768 × 576 pixel) at an exposure time of 4 ms.

Baseline Perfusion and CO₂ Reactivity in Control and STA-MCA Bypass Patients

After surgical exposure, the LSI device was positioned over the imaging field and a 2-minute baseline perfusion measurement was obtained. For CO₂ reactivity assessment, the end-expiratory CO₂ concentration was gradually elevated by 10 mm Hg and the cortical blood flow response was continuously recorded over a time period of 5 to 7 minutes. After the measurement, normoventilation was restored and surgery resumed as planned. In patients undergoing STA-MCA bypass grafting, an end-to-side anastomosis between the frontal or parietal branch of the STA and a cortical M4 segment of the MCA was performed. After confirmation of bypass patency by ICG angiography, LSI was repeated at the same area of the cortex with a second baseline perfusion and CO₂ reactivity measurement.

Bypass Test Occlusion in High- and Intermediate-Flow Bypass Patients

After completion of the high- or intermediate-flow bypass, graft patency was controlled by ICG angiography. The LSI device was positioned over the exposed cortical area and baseline perfusion was continuously recorded. During continuous LSI, the aneurysm-carrying vessel and the bypass were simultaneously clipped for 10 to 12 seconds before flow was initiated through the bypass. If baseline perfusion provided through the bypass was not lower than the perfusion provided by the main vessel, the main vessel was permanently occluded to trap the aneurysm.

Image and Data Processing of Laser Speckle Recordings

Cortical perfusion was recorded with the purpose-designed data acquisition software (MoorFLPI measurement software, Version V3.0, Moor Instruments, Axminster, UK) in the arbitrary perfusion unit CBF-Flux. For the purpose of providing the surgeon with simultaneous real-time visual and (pseudo-)quantitative information on blood flow, the data acquisition software permits a user-defined positioning of up to 16 variable regions of interest (ROIs) within the live speckle image. Within these ROIs, the mean CBF-Flux is calculated in real time and traced as a line graph. For study purposes, CBF-Flux was calculated off-line with the purpose-designed review software (MoorFLPI review software) on the basis of the intraoperatively recorded laser speckle blood flow maps. A user-defined polygonal ROI was placed within the imaging field and the mean cortical CBF-Flux within the ROI was calculated. In cases where the imaging field was partially obstructed (i.e. by a bypass graft), two separate ROIs were placed over the unobstructed cortical area and the mean CBF-Flux was calculated from both ROIs. For repeated measurements before and after STA-MCA bypass grafting, identical ROIs were placed in the same cortical area. Finally, graph traces from all 61 data sets were generated for the time course of each measurement.

In control and STA-MCA bypass patients, baseline perfusion was calculated as the mean CBF-Flux during a 1-minute plateau of the 2-minute baseline perfusion measurement. CO₂ reactivity was expressed as the percent change in CBF-Flux between the baseline plateau and a 1-minute perfusion plateau upon completion of the CO₂ challenge. Graph traces of cortical perfusion during test occlusion and flow initiation in patients undergoing high- or intermediate-flow bypass grafting were expressed as the percent change in perfusion compared with 100% of baseline perfusion.

Statistical Analysis

Data are presented as median and range or mean ± standard deviation. Statistics were generated by GraphPad Prism for Mac (Version 5.0f, GraphPad Software, San Diego, CA, USA). For comparison of physiologic variables (MAP and PaCO₂), baseline perfusion and CO₂ reactivity between control, ACVD, and MMD patients, and for comparison of baseline perfusion before, during, and after test occlusion, a one-way analysis of variance (ANOVA) was performed. For comparison of physiologic variables before and after elevation of PaCO₂ and for comparison of baseline perfusion and CO₂ reactivity before and after STA-MCA bypass grafting, a two-tailed paired t-test was used. Statistical significance was set at P < 0.05.

RESULTS

(Pseudo-)Quantitative Assessment of Relative CBF

We hypothesized that LSI following a fixed protocol should allow (pseudo-)quantitative comparison of CBF-Flux. In our study, LSI permitted robust visualization and measurement of macro- and microcirculatory flow in all control and STA-MCA bypass patients. Mean arterial pressure and PaCO₂ did not differ between groups (Table 2). No difference was found in baseline perfusion between patients with and without hemodynamic compromise (Control: 1994 ± 493 Flux;ACVD: 1936 ± 487 Flux;MMD: 1938 ± 396 Flux (Figure 2B)). In patients who received STA-MCA bypass grafting, LSI detected a significant perisylvian CBF-Flux increase around the site of the anastomosis (ACVD: 2208 ± 631 Flux, ^*P < 0.05 vs before;MMD: 2549 ± 609 Flux, ^*P < 0.01 vs before (Figures 2A and 2C)). The higher relative increase of baseline perfusion after bypass grafting in patients with MMD is illustrated in Figure 2D.

Table 2.

PaCO₂ and MAP before and after elevation of end-expiratory CO₂.

	Before bypass			After bypass
	Control (n = 8)	ACVD (n = 14)	MMD (n = 9)	ACVD (n = 14)	MMD (n = 9)
PaCO₂ ↓ (mm Hg) / MAP (mm Hg)	42 ± 3/85 ± 8	42 ± 3/96 ± 13	40 ± 3 / 95 ± 15	42 ± 4 / 96 ± 12	41 ± 3 / 97± 18
PaCO₂ ↑ (mm Hg) / MAP (mm Hg)	52 ± 5*/82 ± 8	51± 4*/93 ± 12	51 ± 4*/92 ±16	52 ± 3* / 95 ± 11	52 ± 4*/97 ±18

Abbreviations: ACVD, arteriosclerotic cerebrovascular disease; MAP, mean arterial blood pressure; MMD, Moyamoya disease; PaCO₂, arterial partial pressure of carbon dioxide; PaCO₂ ↓, baseline PaCO₂; PaCO₂ ↑, elevated PaCO₂.

Values are expressed as mean ± s.d. ^*P < 0.0001 compared with PaCO₂ ↓.

Figure 2.

(Pseudo-)quantitative assessment of relative cerebral blood flow(CBF). (A) Intraoperative laser speckle imaging screenshots of cortical baseline perfusion before and after superficial temporal artery to middle cerebral artery (STA-MCA) bypass grafting in patients with arteriosclerotic cerebrovascular disease (ACVD; upper panels) and Moyamoya disease (MMD; lower panels). Perfusion is visualized as a two-dimensional color-coded map of blood flow in the arbitrary unit Flux (red = high flow; blue = low flow). The dashed region of interest (ROI) shows the area within which the mean cortical perfusion was calculated. The enhanced red-colored perfusion pattern after bypass completion (right panels, ICG angiography; the asterisk indicates the patent bypass) is an expression of a postoperative perfusion increase after STA-MCA bypass grafting. (B) Bar graph showing the mean cortical baseline perfusion (CBF-Flux) of hemodynamically healthy control patients compared with patients with ACVD and MMD before STA-MCA bypass grafting. (C) Bar graph comparing the mean cortical baseline perfusion (CBF-Flux) before and after STA-MCA bypass grafting in ACVD and MMD patients (^*P < 0.05 for ACVD and ^*P < 0.01 for MMD). (D) Relative (percent) change of baseline perfusion after STA-MCA bypass grafting in 14 patients with ACVD (left) and 9 patients with MMD (right). BP, bypass; CBF, cerebral blood flow.

Applicability and Sensitivity of LSI During Continuous Perfusion Assessment

To evaluate the applicability of intraoperative LSI and characterize the LSI-specific sensitivity for the detection of slight changes in cortical perfusion, we compared CO₂ reactivity in patients with and without hemodynamic impairment, because non-ischemic cerebral hypoperfusion typically implies an impaired vascular reserve. Graph traces of a continuous LSI measurement during elevation of PaCO₂ are shown in Figure 3A. In control patients, elevation of PaCO₂ resulted in a mean cortical perfusion increase of 31.2% ±5%. In patients with SPECT-confirmed hemodynamic compromise, LSI detected a significantly impaired CO₂ reactivity (ACVD: 9.6% ± 9%;MMD: 8.5% ± 8%;^*P < 0.0001 for ACVD and MMD vs control (Figures 3B and 3C)), which was confirmed by a second CO₂ reactivity testing after completion of the bypass graft (ACVD: 9.9% ± 13%;MMD: 2.9% ±5%;^*P < 0.05 for MMD after bypass vs MMD before bypass (Figures 3B and 3D)). The relative decrease of the CO₂ reactivity after bypass grafting in patients with MMD is shown in Figure 3E.

Figure 3.

Applicability and sensitivity of intraoperative laser speckle imaging. (A) Typical cortical perfusion response (CBF-Flux; upper graph traces) and mean arterial blood pressure (MAP in mm Hg; lower graph traces) during a continuous 10mmHg increase of end-expiratory carbon dioxide (CO₂) in a control patient and a patient with arteriosclerotic cerebrovascular disease (ACVD) before and after superficial temporal artery to middle cerebral artery (STA-MCA) bypass grafting. The gray-shaded area illustrates the perfusion change during the CO₂ challenge. (B) Laser speckle images before (upper panels) and after (lower panels) elevation of CO₂ in a control patient (left) and in a patient with ACVD before (center) and after (right) STA-MCA bypass grafting. The dashed region of interest (ROI) shows the area within which the mean cortical perfusion was calculated. (C) Bar graph comparing the mean perfusion change after CO₂ elevation in control, ACVD, and MMD patients before STA-MCA bypass grafting (^*P< 0.0001). (D) Bar graph comparing the mean perfusion change after CO₂ elevation before and after STA-MCA bypass grafting in ACVD and MMD patients (^*P < 0.05). (E) Relative (percent) change of CO₂ reactivity after STA-MCA bypass grafting in 14 patients with ACVD (left) and 9 patients with MMD (right). BP, bypass.

Dynamics of LSI During Acute Perfusion Changes

The temporal response pattern of LSI in regard to acute changes in cortical perfusion was analyzed in a separate group of patients. Laser speckle images of patients undergoing high- or intermediate-flow bypass revascularization are shown in Figure 4A. In all cases, LSI permitted continuous real-time assessment of relative CBF. After test occlusion, LSI visualized and measured an immediate cortical perfusion drop to 56.5% ± 13% of baseline perfusion, which remained at a consistent level for the duration of the occlusion time. After flow initiation through the bypass, cortical perfusion promptly returned to 102.3% ±3% of baseline, indicating a sufficient flow substitution by the bypass (baseline: 3327 ± 569 Flux; occlusion: 1927 ±716 Flux; bypass open: 3408 ± 600 Flux; ^*P < 0.0001 vs occlusion (Figures 4A and 4B)). Individual graph traces of the temporal response pattern and the degree of flow reduction during the test-occlusion procedure are shown in Figure 4C.

Figure 4.

Temporal response properties (dynamics) of laser speckle imaging. (A) Laser speckle imaging screenshots (1 Hz) from a continuous cortical perfusion assessment in a patient after completion of a high-flow bypass graft (saphenous vein; asterisk). From left to right, the panels visualize cortical perfusion before, during, and after a test-occlusion procedure of the bypass and main aneurysm-carrying vessel with subsequent flow initiation only through the bypass graft. The corresponding backscatter photo image is presented on the right. The dashed region of interest (ROI) shows the area within which the mean cortical perfusion was calculated. The LSI-specific detection of a sudden flow reduction during test occlusion and immediate reperfusion after flow initiation can be noted in the second and third panels from the left. Note the absent perfusion in the bypass graft during test occlusion (asterisk). (B) Bar graph comparing the mean cortical perfusion (CBF-Flux) of all patients during test occlusion and flow initiation (^*P < 0.0001). (C) Individual graph traces of cortical perfusion (percent change of cortical perfusion) in patients undergoing high- or intermediate-flow bypass grafting. t, temporal; f, frontal.

Graft Patency Assessment Through LSI

Laser speckle imaging visualized positive perfusion within all STA-MCA, high-, and intermediate-flow bypass grafts (30/30), which was later confirmed by ICG angiography. In patients undergoing high- or intermediate-flow bypass grafting, visual assessment of a patent or non-patent bypass graft corresponded with cortical hypo- or normoperfusion during test occlusion and flow initiation, respectively.

DISCUSSION

In this study, we demonstrated the general applicability of LSI for continuous, non-invasive real-time imaging of relative CBF during a series of standardized CBF variations in patients undergoing neurovascular procedures. Laser speckle imaging proved to be easy-to-use, robust and was characterized by excellent image quality in high spatial resolution. All patients tolerated the procedure well and side effects were not observed. Repeated measurements did not pose any problems in regard to patient safety and image quality. Under standardized conditions, LSI permitted quantitative assessment of relative CBF. The system featured a high temporal resolution and a high sensitivity to even minor flow changes, which is mandatory for monitoring CBF under physiologic and pathologic conditions.

Intraoperative Imaging of CBF

Intraoperative functional diagnostics have increased neurosurgical efficiency and patient safety.^13-15 However, the key determinant of tissue viability remains CBF, which to date cannot be adequately assessed intraoperatively. Although numerous imaging modalities, such as positron emission tomography (PET), SPECT, xenon-enhanced CT, dynamic perfusion CT, or magnetic resonance imaging (MRI) perfusion studies are used to study cerebral hemodynamics, intraoperative application of this technology remains limited by immobility, technical requirements, long data acquisition and processing times.

Non-invasive optical imaging has the advantage of reduced cost and infrastructure requirements, which facilitate intraoperative application. In general, the ideal technique should be sensitive enough to detect harmful events, quantitative, continuous, non-invasive, robust, user-friendly, and cost-effective and have a high spatial-temporal resolution. However, none of the available optical imaging techniques for intraoperative assessment of cerebral perfusion meet these criteria.^3,6,16–19 The current gold standard for intraoperative visualization of perfusion within the vasculature is ICG angiography.^3,12,20 Although the spatial-temporal resolution is high, the methodology requires injection of an exogenous substance and does not permit direct and continuous real-time assessment of relative CBF. For this purpose, LSI serves as an optical blood flow visualization and measurement tool based on dynamic light scattering off moving particles (erythrocytes).^7,21 In this study, we evaluated the general applicability of intraoperative LSI because, at present, no data exist on direct cortical blood flow assessment in the human brain and information on procedure-related blood flow changes is of leading interest during neurosurgery. Moreover, robustness, comparability, and (pseudo-)quantitative assessment of relative CBF by LSI must be addressed in the clinical setting to allow meaningful data interpretation.

(Pseudo-)Quantitative Assessment of Relative CBF

Arbitrary blood flow velocity measurements by LSI are naturally influenced by confounding factors such as the distance and angle of the laser to the imaging surface, the size of the imaging field, ambient lighting, polarizing reflexes, the exposure time and the temporal resolution of image acquisition, among others. Against this background, recent evidence suggests that under standardized conditions, LSI may provide (pseudo-) quantitative information on flow.^21-24 In our study, standardized patient management, device setup and system configuration permitted robust assessment of relative CBF and our findings were in line with the normal or at best mild reduction in resting CBF that is typically noted in patients with hemodynamic impairment.

After completion of the STA-MCA bypass, LSI detected a significant perfusion increase in the area surrounding the bypass anastomosis. This phenomenon has previously been reported after flow augmentation in patients with hemodynamic compromise,²⁵ where the chronic critical reduction in cerebral perfusion may induce a persistent maximal dilation of the cerebral arterioles, causing postoperative hyperperfusion in response to the rapid perfusion recovery, which may or may not be associated with transient neurologic deficits. In case of flow augmentation through STA-MCA bypass grafting, the postoperative perfusion increase is typically confined to the perisylvian area around the bypass anastomosis and the clinical presentation of symptomatic hyperperfusion may mimic that of ischemic stroke with sudden onset, upper extremity (hand) motor dysfunction, motor or sensory aphasia, dysarthria, and seizure activity.^26-28 Therefore, a tool for early and easy distinction of hyperperfusion from ischemic stroke is desirable. At present, the only positive risk factors for development of symptomatic hyperperfusion after STA-MCA bypass grafting are reported as (1) a preoperatively increased oxygen extraction fraction²⁷ and (2) a cerebral blood volume increase in adults.²⁸ However, routine assessment of CBF and metabolism via PET poses high logistical and financial challenges and is not readily available. For this purpose, LSI could serve as a noninvasive, cost-effective, and easy-to-use method to analyze the local perfusion change after successful STA-MCA bypass grafting. Controlled and prospective studies with larger patient numbers for comparison of LSI-specific perfusion changes with quantitative CBF measurements are needed for further validation of the methodology and could help establish precision rates (positive and negative predictive values) of perfusion thresholds; phase-specific treatment strategies could further reduce surgical complications and improve long-term outcome.

Applicability and Sensitivity

The applicability and sensitivity of intraoperative LSI was characterized by analyzing standardized LSI-specific responses to PaCO₂ elevation because consistent results have been reported on cerebral vasodilatation to hypercapnia^29,30 and CO₂ reactivity is impaired in patients with hemodynamic compromise.³¹ The CBF increase of 31% after a PaCO₂ increase of 10 mm Hg in the control group of our study was in line with the 30% to 40% increase of CBF per 10 mm Hg increase of PaCO₂ in patients undergoing general anesthesia with propofol and remifentanil.^17,32 Although propofol has been shown to lower the cerebral metabolic rate, neuronal activity, cerebral blood volume and CBF,^33,34 CO₂ reactivity is considered to remain intact compared with awake patients.^{32,33,35–37} Because propofol and remifentanil are the anesthetic agents of choice for intracranial surgery,³³ intraoperative CO₂ reactivity testing could therefore permit a more robust characterization of perfusion abnormalities compared with baseline perfusion measurements alone. Further assessment of cerebrovascular CO₂ reactivity is relevant because baseline CBF measurements alone may not sufficiently predict the risk of ischemia, as demonstrated in patients with hemodynamic compromise.³⁸

LSI sensitively detected cortical perfusion changes during continuous assessment of relative CBF by registering a limited CO₂ reactivity in patients with hemodynamic impairment. Although we did not expect an immediate recovery of the CO₂ reactivity after STA-MCA bypass grafting, further reduction of the CO₂ reactivity in patients with MMD was an unexpected finding. The following hypothesis could explain these results: First, cerebrovascular reactivity after STA-MCA bypass grafting may not immediately improve, because changes in PaCO₂ do not regulate the net flow though the STA donor vessel. Second, the fixed dilation of the cerebral resistance vessels in chronic cerebral hypoperfusion could require some time to adjust to the new hemodynamic situation and may hamper autoregulatory vasoconstriction after external flow augmentation. This hypothesis is in line with the postoperative perfusion increase in ACVD and MMD patients, despite a similar MAP before and after completion of the bypass graft. In addition, surgical manipulation could lead to a greater postoperative impairment of CO₂ reactivity in patients with MMD due to the altered morphologic structure of the vascular wall³⁹ and increased microvascular density in these patients.⁴

Temporal Dynamics

A suitable technology for continuous intraoperative CBF assessment must allow real-time detection of acute perfusion changes. LSI blood flow maps should be easy to interpret in a way such that brief visual inspection provides useful information in regard to tissue integrity and perfusion. In our experience, this was given by implementation of a temporal filter, which reduced the temporal resolution but enhanced the spatial detail resolution and improved the signal-to-noise ratio. To test whether the response properties of the system were still suitable for real-time perfusion assessment after implementation of the temporal filter, we tested the dynamics of the system during acute changes in cortical perfusion. Our results show that a scan rate of 1 Hz permitted detection of acute perfusion changes in a dynamic fashion while visualizing cortical perfusion in excellent detail resolution. The temporal filter additionally reduced the measurement variability associated with motion artifacts due to pulsation of the brain and facilitated comparison of the LSI and the macroscopic view of the investigated area. While it was not our intention to evaluate LSI as a tool for intraoperative functional testing, this remains an approach that could be addressed in future studies. For this purpose, however, the scan rates that we selected would be too low and could be increased to 25 Hz in order to maximize temporal resolution.

Bypass Graft Patency Assessment

Intraoperative monitoring of cerebral perfusion in the macro- and microvasculature requires selection of the most appropriate technique. For the purpose of visualizing and measuring relative CBF in the living human brain, we demonstrated the validity and applicability of LSI in a routine setting. For the purpose of graft patency assessment, however, it should be noted that LSI was limited by an obstruction of the imaging field (particularly through the adventitial cuff surrounding an STA-MCA bypass graft) and a lower spatial detail resolution compared with ICG angiography. Although it was ultimately possible to confirm positive flow within at least some part of the bypass graft in all cases, LSI did not permit interpretation of flow direction, vessel stenosis or patency of the bypass anastomosis. For the purpose of perfusion visualization within the vasculature, ICG angiography should therefore remain the method of choice.

Limitations

A main technical limitation of the presented methodology is that LSI can only assess blood flow in superficial vascular territories that are exposed surgically. Further, the system we used is currently not integrated into the surgical microscope. The clear benefits of microscope integration are its greater ease-of-use and better incorporation into the surgical workflow. To minimize this disadvantage, a separate surgeon performed assembly of the system without interrupting the surgical workflow and mostly the system remained assembled between measurements. The same surgeon was also responsible for operating the LSI device during each measurement, which increased the standardization and robustness of the procedure. As an advantage, the versatility of our setup permitted its use in different operating theaters for the purpose of back-to-back measurements in different patients.

DISCLOSURE/CONFLICT OF INTEREST

The authors declare no conflict of interest.

References

Albers

Thijs

Wechsler

Kemp

Schlaug

Skalabrin

Magnetic resonance imaging profiles predict clinical response to early reperfusion: the diffusion and perfusion imaging evaluation for understanding stroke evolution (DEFUSE) study. Ann Neurol 2006; 60: 508–517.

Khatri

Abruzzo

Yeatts

Nichols

Broderick

Tomsick

Good clinical outcome after ischemic stroke with successful revascularization is time-dependent. Neurology 2009; 73: 1066–1072.

Raabe

Beck

Gerlach

Zimmermann

Seifert

. Near-infrared indocyanine green video angiography: a new method for intraoperative assessment of vascular flow. Neurosurgery 2003; 52: 132–139 discussion 139.

Czabanka

Pena-Tapia

Schubert

Woitzik

Vajkoczy

Schmiedek

. Characterization of cortical microvascularization in adult moyamoya disease. Stroke 2008; 39: 1703–1709.

Woitzik

Pena-Tapia

Schneider

Vajkoczy

Thome

. Cortical perfusion measurement by indocyanine-green videoangiography in patients undergoing hemicraniectomy for malignant stroke. Stroke 2006; 37: 1549–1551.

Kamp

Slotty

Turowski

Etminan

Steiger

Hanggi

Microscope-integrated quantitative analysis of intraoperative indocyanine green fluorescence angiography for blood flow assessment: first experience in 30 patients. Neurosurgery 2012; 70(1 Suppl Operative): 65–73(discussion 73–74).

Briers

. Laser Doppler, speckle and related techniques for blood perfusion mapping and imaging. Physiol Meas 2001; 22: R35–R66.

Briers

Fercher

. Retinal blood-flow visualization by means of laser speckle photography. Invest Ophthalmol Vis Sci 1982; 22: 255–259.

Dunn

Bolay

Moskowitz

Boas

. Dynamic imaging of cerebral blood flow using laser speckle. J Cereb Blood Flow Metab 2001; 21: 195–201.

10.

Hecht

Woitzik

Dreier

Vajkoczy

. Intraoperative monitoring of cerebral blood flow by laser speckle contrast analysis. Neurosurg Focus 2009; 27: E11.

11.

Parthasarathy

Weber

Richards

Fox

Dunn

. Laser speckle contrast imaging of cerebral blood flow in humans during neurosurgery: a pilot clinical study. J Biomed Opt 2010; 15: 066030.

12.

Woitzik

Horn

Vajkoczy

Schmiedek

. Intraoperative control of extracranial-intracranial bypass patency by near-infrared indocyanine green videoangiography. J Neurosurg 2005; 102: 692–698.

13.

Berman

Berger

Mukherjee

Henry

. Diffusion-tensor imaging-guided tracking of fibers of the pyramidal tract combined with intraoperative cortical stimulation mapping in patients with gliomas. J Neurosurg 2004; 101: 66–72.

14.

Cannestra

Bookheimer

Pouratian

O'Farrell

Sicotte

Martin

Temporal and topographical characterization of language cortices using intraoperative optical intrinsic signals. Neuroimage 2000; 12: 41–54.

15.

Duffau

. A personal consecutive series of surgically treated 51 cases of insular WHO Grade II glioma: advances and limitations. J Neurosurg 2009; 110: 696–708.

16.

Bolognese

Miller

Heger

Milhorat

. Laser-Doppler flowmetry in neurosurgery. J Neurosurg Anesthesiol 1993; 5: 151–158.

17.

Klein

Schramm

Glaser

Reisch

Tresch

Werner

Intraoperative monitoring of cerebral microcirculation and oxygenation-a feasibility study using a novel photo-spectrometric laser-Doppler flowmetry. J Neurosurg Anesthesiol 2010; 22: 38–45.

18.

Nakagawa

Fujimura

Arafune

Sakuma

Tominaga

. 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: 1158–1164.

19.

Raabe

Van De Ville

Leutenegger

Szelenyi

Hattingen

Gerlach

Laser Doppler imaging for intraoperative human brain mapping. Neuroimage 2009; 44: 1284–1289.

20.

Hanggi

Etminan

Steiger

. The impact of microscope-integrated intraoperative near-infrared indocyanine green videoangiography on surgery of arteriovenous malformations and dural arteriovenous fistulae. Neurosurgery 2010; 67: 1094–1103(discussion 1103–1104).

21.

Boas

Dunn

. Laser speckle contrast imaging in biomedical optics. J Biomed Opt 2010; 15: 011109.

22.

Duncan

Kirkpatrick

. Can laser speckle flowmetry be made a quantitative tool? J Opt Soc Am A Opt Image Sci Vis 2008; 25: 2088–2094.

23.

Jones

Shin

Boas

Hyman

Moskowitz

Ayata

Simultaneous multispectral reflectance imaging and laser speckle flowmetry of cerebral blood flow and oxygen metabolism in focal cerebral ischemia. J Biomed Opt 2008; 13: 044007.

24.

Strong

Bezzina

Anderson

Boutelle

Hopwood

Dunn

. Evaluation of laser speckle flowmetry for imaging cortical perfusion in experimental stroke studies: quantitation of perfusion and detection of peri-infarct depolarisations. J Cereb Blood Flow Metab 2006; 26: 645–653.

25.

Piepgras

Morgan

Sundt

TMJ

Yanagihara

Mussman

. Intracerebral hemorrhage after carotid endarterectomy. J Neurosurg 1988; 68: 532–536.

26.

Fujimura

Mugikura

Kaneta

Shimizu

Tominaga

. Incidence and risk factors for symptomatic cerebral hyperperfusion after superficial temporal artery-middle cerebral artery anastomosis in patients with moyamoya disease. Surg Neurol 2009; 71: 442–447.

27.

Kaku

Iihara

Nakajima

Kataoka

Fukuda

Masuoka

Cerebral blood flow and metabolism of hyperperfusion after cerebral revascularization in patients with moyamoya disease. J Cereb Blood Flow Metab 2012; 32: 2066–2075.

28.

Uchino

Kuroda

Hirata

Shiga

Houkin

Tamaki

. Predictors and clinical features of postoperative hyperperfusion after surgical revascularization for moyamoya disease: a serial single photon emission CT/positron emission tomography study. Stroke 2012; 43: 2610–2616.

29.

Kastrup

Thomas

Hartmann

Schabet

. Sex dependency of cerebrovascular CO2 reactivity in normal subjects. Stroke 1997; 28: 2353–2356.

30.

Kety

Schmidt

. The effects of active and passive hyperventilation on cerebral blood flow, cerebral oxygen consumption, cardiac output, and blood pressure of normal young men. J Clin Invest 1946; 25: 107–119.

31.

Conklin

Fierstra

Crawley

Han

Poublanc

Mandell

Impaired cerebrovascular reactivity with steal phenomenon is associated with increased diffusion in white matter of patients with Moyamoya disease. Stroke 2010; 41: 1610–1616.

32.

Dagal

Lam

. Cerebral autoregulation and anesthesia. Curr Opin Anaesthesiol 2009; 22: 547–552.

33.

Cole

Gottfried

Gupta

Couldwell

. Total intravenous anesthesia: advantages for intracranial surgery. Neurosurgery 2007; 61(Suppl 2): 369–377 (discussion 377–8).

34.

Kaisti

Langsjo

Aalto

Oikonen

Sipila

Teras

Effects of sevoflurane, propofol, and adjunct nitrous oxide on regional cerebral blood flow, oxygen consumption, and blood volume in humans. Anesthesiology 2003 2003; 99: 603–613.

35.

Conti

Iacopino

Fodale

Micalizzi

Penna

Santamaria

. Cerebral haemodynamic changes during propofol-remifentanil or sevoflurane anaesthesia: transcranial Doppler study under bispectral index monitoring. Br J Anaesth 2006; 97: 333–339.

36.

Fox

Gelb

Enns

Murkin

Farrar

Manninen

. The responsiveness of cerebral blood flow to changes in arterial carbon dioxide is maintained during propofolnitrous oxide anesthesia in humans. Anesthesiology 1992; 77: 453–456.

37.

Harrison

Girling

Mahajan

. Effects of target-controlled infusion of propofol on the transient hyperaemic response and carbon dioxide reactivity in the middle cerebral artery. Br J Anaesth 1999; 83: 839–844.

38.

Yonas

Smith

Durham

Pentheny

Johnson

. Increased stroke risk predicted by compromised cerebral blood flow reactivity. J Neurosurg 1993; 79: 483–489.

39.

Lin

Xie

Zhang

Tan

Clinical and immunopathological features of Moyamoya disease. PLoS One 2012; 7: e36386.

Laser Speckle Imaging Allows Real-Time Intraoperative Blood Flow Assessment During Neurosurgical Procedures

Abstract

Keywords

Introduction

MATERIALS AND METHODS

Study Design

Patient Management

Technical Specifications

Device Setup

System Configuration

Baseline Perfusion and CO2 Reactivity in Control and STA-MCA Bypass Patients

Bypass Test Occlusion in High- and Intermediate-Flow Bypass Patients

Image and Data Processing of Laser Speckle Recordings

Statistical Analysis

RESULTS

(Pseudo-)Quantitative Assessment of Relative CBF

Applicability and Sensitivity of LSI During Continuous Perfusion Assessment

Dynamics of LSI During Acute Perfusion Changes

Graft Patency Assessment Through LSI

DISCUSSION

Intraoperative Imaging of CBF

(Pseudo-)Quantitative Assessment of Relative CBF

Applicability and Sensitivity

Temporal Dynamics

Bypass Graft Patency Assessment

Limitations

DISCLOSURE/CONFLICT OF INTEREST

References

Baseline Perfusion and CO₂ Reactivity in Control and STA-MCA Bypass Patients