Intraoperative Cerebral Hemodynamic Monitoring during Carotid Endarterectomy via Diffuse Correlation Spectroscopy and Near-Infrared Spectroscopy

Abstract

Objective: This pilot study aims to show the feasibility of noninvasive and real-time cerebral hemodynamic monitoring during carotid endarterectomy (CEA) via diffuse correlation spectroscopy (DCS) and near-infrared spectroscopy (NIRS). Methods: Cerebral blood flow index (CBFi) was measured unilaterally in seven patients and bilaterally in seventeen patients via DCS. In fourteen patients, hemoglobin oxygenation changes were measured bilaterally and simultaneously via NIRS. Cerebral autoregulation (CAR) and cerebrovascular resistance (CVR) were estimated using CBFi and arterial blood pressure data. Further, compensatory responses to the ipsilateral hemisphere were investigated at different contralateral stenosis levels. Results: Clamping of carotid arteries caused a sharp increase of CVR (~70%) and a marked decrease of ipsilateral CBFi (57%). From the initial drop, we observed partial recovery in CBFi, an increase of blood volume, and a reduction in CVR in the ipsilateral hemisphere. There were no significant changes in compensatory responses between different contralateral stenosis levels as CAR was intact in both hemispheres throughout the CEA phase. A comparison between hemispheric CBFi showed lower ipsilateral levels during the CEA and post-CEA phases (p < 0.001, 0.03). Conclusion: DCS alone or combined with NIRS is a useful monitoring technique for real-time assessment of cerebral hemodynamic changes and allows individualized strategies to improve cerebral perfusion during CEA by identifying different hemodynamic metrics.

Keywords: carotid endarterectomy; cerebral autoregulation; cerebral blood flow; diffuse correlation spectroscopy; intraoperative neuromonitoring; near-infrared spectroscopy.

Conflict of interest statement

M.A.F. has a financial interest in 149 Medical, Inc., a company developing DCS technology for assessing and monitoring cerebral blood flow in newborn infants. M.A.F. interests were reviewed and managed by Massachusetts General Hospital and Mass General Brigham according to their conflict-of-interest policies. All other authors reported no conflicts of interest.

Figures

Figure 1

(A) The DCS-NIRS probe. The co-localized DCS and NIRS source fiber is marked by the red square. The co-localized DCS and NIRS short separation detector fiber (blue square) is located at 5 mm from the source. The short separation detector is employed to measure superficial (extracerebral) blood flow and hemoglobin oxygenation changes. A DCS long separation detector (orange square) is located at 25 mm (it includes three detectors to improve SNR), and a NIRS long separation detector (green square) is located at 30 mm from the source. Long separations are employed to measure cerebral blood flow and hemoglobin oxygenation. (B) The optical probes after bilateral placement on one patient in conjunction with EEG electrodes. (C) Schematic timeline of the surgery and data analysis phases. The signal average of a physiologically stable 3-min period immediately before clamping was used to normalize the data to quantify the transient ipsilateral changes of clamping and unclamping across patients (red shaded area, short baseline). The pre-CEA phase (gray and red shaded areas, long baseline) was used to normalize the data to quantify changes in the CEA (green shaded area) and post-CEA (blue shaded area) phases with respect to the pre-CEA phase. The 3 min of transient data between the two phases were discarded from the second analysis.

Figure 2

A typical measurement on a patient (#21, MCS group) during CEA. Ipsilateral and contralateral cerebral hemodynamic changes are shown in blue and orange, respectively, for rCBFi (A), rCVR (B), ∆HbT (C), and ∆SO2 (D). MAP ((E), black) was measured invasively via an arterial cannula in the arm. All data are normalized with respect to the pre-CEA phase (long baseline), represented in grey and red shaded areas, except MAP. A three-minute short baseline is represented in the red shaded area. CEA and post-CEA phases are represented in shaded light green and blue areas, respectively. CEA, carotid endarterectomy; rCBFi, relative cerebral blood flow index; rCVR, relative cerebrovascular resistance; ∆SO2, changes in oxygen saturation; ∆HbT, changes in total hemoglobin concentration; MAP, mean arterial pressure.

Figure 3

A quantified summary of short-term transient changes in clamping and unclamping in each group. rCBFi, ∆HbT, and ∆SO2 were quantified with respect to the short baseline. All data are represented in a whisker plot. Black-filled circles represent outliers. rCBFi, relative cerebral blood flow index; ∆HbT, changes in total hemoglobin concentration; ∆SO2, changes in oxygen saturation; MCS, normal-to-mild contralateral stenosis group (<50%); SCS, moderate-to-substantial contralateral stenosis group (≥50%); SH, shunt utilization group.

Figure 4

Hemodynamic changes in rCBFi, rCVR, and MAP of all patients during CEA and post-CEA phases. All changes were normalized with respect to the pre-CEA phase (long baseline, gray dashed line). All data are represented in mean ± standard error of the mean. Changes in the ipsilateral hemisphere were statistically significant than in the contralateral hemisphere both during CEA (indicated as *) and post-CEA phases (indicated as #). CEA, carotid endarterectomy; rCBFi, relative cerebral blood flow index; rCVR, relative cerebrovascular resistance; MAP, mean arterial pressure.

Figure 5

A representative case of ipsilateral (A) and contralateral (B) CBFi versus MAP in a patient (#21, MCS group). In the ipsilateral hemisphere, there was a decrease of CBFi for MAP below 85 mmHg during the post-CEA phase, suggesting the loss of cerebral autoregulation for these MAP values. In contrast, during the pre-CEA and CEA phases, CBFi did not increase despite MAP being above 120 mmHg, as this patient was still in the autoregulatory range. The pre-CEA, CEA, and post-CEA phases are represented as grey diamond, green dotted, and blue triangle symbols, respectively. CAR values are the computed cross-correlation coefficients with respect to our method. The least-squares method is used to find a polynomial fit of the maximum degree of two and R2 values are the results of curve fitting. The fit results are shown in color with respect to phases. CBFi, cerebral blood flow index; MAP, mean arterial pressure; CEA, carotid endarterectomy; CAR, cerebral autoregulation.

Figure 6

Mean correlation coefficients (r) between CBFi and MAP during pre-CEA, CEA, and post-CEA phases in the MCS, SCS, and SH groups. The ipsilateral hemisphere is blue, and the contralateral hemisphere is represented in orange color. CBFi was more autoregulated in the SCS group than MCS and SH groups during the pre-CEA phase (indicated as #, p = 0.003). The intact CAR threshold (r = 0.3) is indicated as a red dashed line. Data are shown as mean ± standard error of the mean. * p = 0.03. CAR, cerebral autoregulation; CBFi, cerebral blood flow index; MAP, mean arterial pressure; CEA, carotid endarterectomy; MCS, mild contralateral stenosis group (<50%); SCS, substantial contralateral stenosis group (≥50%); SH, shunt group.