Phase-dependent Brain Activation of the Frontal and Parietal Regions During Walking After Stroke - An fNIRS Study

Abstract

Background: Recovery of walking post-stroke is highly variable. Accurately measuring and documenting functional brain activation characteristics during walking can help guide rehabilitation. Previous work in this area has been limited to investigations of frontal brain regions and have not utilized recent technological and analytical advances for more accurate measurements. There were three aims for this study: to characterize the hemodynamic profile during walking post-stroke, to investigate regional changes in brain activation during different phases of walking, and to related brain changes to clinical measures.

Methods: Functional near-infrared spectroscopy (fNIRS) along the pre-frontal, premotor, sensorimotor, and posterior parietal cortices was used on twenty individuals greater than six months post-stroke. Individual fNIRS optodes were digitized and used to estimate channel locations on each participant and short separation channels were used to control for extracerebral hemodynamic changes. Participants walked at their comfortable pace several times along a hallway while brain activation was recorded. Exploratory cluster analysis was conducted to determine if there was a link between brain activation and clinical measures.

Results: Sustained activation was observed in the pre-frontal cortex with the ipsilesional hemisphere showing greater activation compared to the contralesional side. Sensorimotor cortex was active during the early, acceleration stage of walking only. Posterior parietal cortex showed changes in activation during the later, steady-state stage of walking. Faster gait speeds also related to increased activation in contralesional sensorimotor and posterior parietal cortices. Exploratory analysis clustered participants into two distinct groups based on their brain activation profiles and generally showed that individuals with greater activation tended to have better physical outcomes.

Conclusions: These findings can guide future research for obtaining adequate power and determining factors that can be used as effect modifiers to reduce inter-subject variability. Overall, this is the first study to report specific oxygenated and deoxygenated hemoglobin changes in frontal to parietal regions during walking in the stroke population. Our results shed light on the importance of measuring brain activation across the cortex and show the importance of pre-frontal, sensorimotor, and posterior parietal cortices in walking after a stroke.

Keywords: functional near-infrared spectroscopy (fNIRS); gait; posterior parietal cortex (PPC); pre-frontal cortex (PFC); premotor cortex (PMC); sensorimotor cortex (SMC); stroke.

Figure 1

(A). Schematic of optode configuration and (B). Example of a participant wearing the fNIRS device.

Figure 2

Average Hemodynamic Responses with Normal–paced Walking for each region of interest. Thick lines indicate group mean response and thin lines indicates 2 standard errors around the mean.

Figure 3

Average (A) HbO and (B) HbR response for each time segment with for each region of interest. Bars indicate average responses for the contralesional side (white) and ipsilesional side (colored) and error lines indicate standard error. ^#Indicates a differences from Baseline with alpha at 0.05 *Indicates a significant difference from Baseline after correction for multiple comparisons.

Figure 4

Relationships between HbO and impairment, performance, and automaticity for each ROI Ipsilesional and Contralesional hemisphere are plotted on a single graph and indicated by red and gray markers; respectively. Additionally; individuals in Cluster 1 are denoted by a triangle marker and individuals in Cluster 2 are denoted by circle markers. ^#Indicates a relationship with alpha at 0.05.