Date of Award

Fall 2012

Document Type


Degree Name

Doctor of Philosophy (PhD)


Biomedical Engineering

First Advisor

Schindler-Ivens, Sheila M.

Second Advisor

Schmit, Brian D.

Third Advisor

Ropella, Kristina M.


This study aimed to enhance our understanding of supraspinal control of locomotion in stroke survivors and its relationship to locomotor impairment. We focused mainly on the locomotor component of walking, which involves rhythmic, reciprocal, flexion and extension movements of multiple joints in both legs. Functional magnetic resonance imaging (fMRI) was used to record human brain activity while pedaling was used as a model of locomotion. First, we examined the spatiotemporal characteristics of hemodynamic responses recorded with fMRI and found that they were different in stroke survivors and control subjects. However, these differences were not substantial enough to require altering the normal canonical hemodynamic response function to obtain valid measurements of pedaling-related brain activity. During pedaling, stroke survivors and control subjects showed activity in the sensorimotor cortex and cerebellum. Stroke survivors had reduced volume of activation in those regions, however the signal intensity was similar between the groups. In stroke survivors, sensorimotor cortex activity was symmetrically distributed across the damaged and undamaged hemispheres; while cerebellum activity was lateralized to the damaged hemisphere. These brain activation patterns were different from those observed during non-locomotor movements, where volume of activation was unchanged but signal intensity was reduced in stroke survivors. We conclude that neural adaptations for producing locomotor and non-locomotor movements post-stroke are not the same and that the spinal cord and cerebellum might have a compensatory role in producing hemiparetic locomotion. Finally, we examined the relationship between locomotor performance and pedaling-related brain activity measured with fMRI. We found no relationship between the brain activation symmetry and locomotor symmetry, suggesting that the brain activation from each hemisphere was not directly responsible for control of the contralateral leg. However, our stroke survivors demonstrated poor locomotor performance and decreased volume of activation measured during pedaling, suggesting that impaired locomotion was associated with reduced volume of activation. Signal intensity of brain activity was associated with rate of pedaling in stroke survivors, suggesting that increased signal intensity in the active brain areas may compensate for reduced volume of activation in the production of hemiparetic locomotion.