Quantification of the Tissue Oxygenation Delay Induced by Breath-Holding in Patients with Carotid Atherosclerosis

Figures

• 

  Figure 1. fNIRS Data acquisition. (a) Brain regions monitored with fNIRS, covering the main brain areas irrigated by the carotid artery. The color bar represents a map of photon sensitivity based on Monte Carlo simulations performed with AtlasViewer [31]; the red color represents regions with high photon sensitivity, while blue represents low–photon sensitivity. (b) Experimental protocol performed by all participants, consisting of up to seven breath-holding trials of 10 s each, followed by 30 s of rest for a total of 6 min of acquisition.

• 

  Figure 2. Hemodynamic response function (HRF) models used in the proposed adaptive GLM (aGLM) analysis. By varying the dispersion time of the HRF basis function (τh in Equation (1)), we could model differences in the transit time of the cerebrovascular response to breath−holding in patients diagnosed with carotid artery stenosis (CAS). The HRFs with different delays (solid curves) were obtained by convolving the HRF basis function with a boxcar representing the breath-holding challenge (BH). The resulting HRFs were characterized by their delayed onset, ranging from 1 s to 7 s.

• 

  Figure 3. Group averaged time-courses of the oxy−hemoglobin (HbO, red), deoxy−hemoglobin (HbR, blue), and total hemoglobin (HbT, green) concentration changes due to the breath-holding task performed by (a) control, (b) unilateral, and (c) bilateral groups. The solid curves represent the median values across all subjects in each group, while the shadows around the curves represent the standard error of the mean. The gray shadow in the plots represents the task duration.

• 

  Figure 4. Group averaged time courses for each hemisphere due to the breath−holding task. The plots show the changes in oxy−hemoglobin (HbO, red), deoxy−hemoglobin (HbR, blue), and total hemoglobin (HbT, green) induced by breath−holding for each group. The left column represents the left hemisphere (controls) or the hemisphere ipsilateral to either the nonoccluded ICA (unilateral patients) or the ICA with the lowest level of occlusion (bilateral patients). The right column shows the data from the right hemisphere (controls) or the hemisphere ipsilateral to either the stenotic/occluded ICA (unilateral patients) or the ICA with the highest level of occlusion (bilateral patients). The solid curves represent the median values across subjects, while the color shadows represent the standard error of the mean. The gray shadow in the plots represents the task duration.

• 

  Figure 5. Individual responses to the breath-holding challenge from four arbitrary subjects within each group. Each sphere on the head represents an fNIRS channel that showed a significant vasodilatory response (as measured by a significant increase in HbT) following apnea. The color bar represents the time delays of the modeled HRF that maximizes the fit of the measured time courses.

• 

  Figure 6. Distribution of the (a) fraction of the activated brain channels, (b) average time delay across the different brain regions, and (c) dispersion of the time delay across the different brain regions as a function of carotid artery stenosis (CAS) severity. Gray: Control, green: unilateral level 1, blue: unilateral level 2, yellow: bilateral level 3 and, orange: bilateral level 4.

• 

  Figure 7. Distribution of the laterality index (LI) as a function of the carotid artery stenosis (CAS) calculated for (a) the fraction of active channels that responded to the breath−holding challenge and the (b) time delays of the vasodilatory hemodynamic response to breath−holding. Gray: Control, green: unilateral level 1, blue: unilateral level 2, yellow: bilateral level 3 and, orange: bilateral level 4.