Date of Award

Spring 2012

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Biomedical Engineering - (Ph.D.)

Department

Biomedical Engineering

First Advisor

Sergei Adamovich

Second Advisor

Eugene Tunik

Third Advisor

Alma S. Merians

Fourth Advisor

Richard A. Foulds

Fifth Advisor

Jason Steffener

Abstract

This dissertation examines manipulation of visual feedback in virtual reality (VR) to increase excitability of distinct neural networks in the sensorimotor cortex. The objective is to explore neural responses to visual feedback of motor activities performed in complex virtual environments during functional magnetic resonance imaging (fMRI), and to identify sensory manipulations that could further optimize VR rehabilitation of persons with hemiparesis. In addition, the effects of VR therapy on brain reorganization are investigated. An MRI-compatible VR system is used to provide subjects with online visual feedback of their hand movement. First, the author develops a protocol to analyze variability in movement kinematics between experimental sessions and conditions and its possible effect on modulating neural activity. Second, brain reorganization after 2 weeks of robot-assisted VR therapy is examined in 10 chronic stroke subjects in terms of change in extent of activation, interhemispheric dominance, connectivity network of ipsilesional primary motor cortex (iM1) and the interhemispheric interaction between iM 1 and contralesional M1 (cM 1). After training, brain activity during a simple paretic hand movement is re-localized in terms of bilateral change in activity or a shift of interhemispheric dominance (re-lateralization) toward the ipsilesional hemisphere that is positively correlated with improvement in clinical scores. Dynamic causal modeling (DCM) shows that interhemispheric coupling between the bilateral motor cortices tends to decrease after training and to negatively correlate with improvement in scores for clinical scales, and with the amount of re-lateralization. Third, the dissertation studies if visual discordance in VR of finger movement would facilitate activity in select brain networks. In a study of 12 healthy subjects, the amplitude of finger movement is manipulated (hypometric feedback) resulting in higher activation of contralateral M1. In a group of 11 stroke subjects, bidirectional, hypometric and hypermetric,VR visual discordance is used. Both feedback conditions cause small increase in activity of the iM1 contralateral to movement and stronger recruitment of both posterior parietal cortices and the ipsilesional fusiform gyrus (iFBA). Fourth, the effect of mirrored-visual feedback on the activity of the ipsilesional sensorimotor cortex of stroke subjects is examined. While subjects move the non-paretic hand during the fMRI experiment, they receive either veridical feedback of the movement or a mirrored feedback. The results show recruitment of iM1 and both posterior parietal cortices during the mirrored feedback. Effective connectivity analysis show increase correlation of iM1 and contralesional SPL (cSPL) with iFBA suggesting a role of the latter in the evaluation of feedback and in visuomotor processing. DCM analysis shows increased modulation of iM1 by cSPL area during the mirrored feedback, an observation that proves the influence of visual feedback on modulating primary motor cortex activation. This dissertation provides evidence that it is possible to enhance brain excitability through manipulation of virtual reality feedback and that brain reorganization can result from just two weeks of VR training. These findings should be exploited in the design of neuroscience-based rehabilitation protocols that could enhance brain reorganization and motor recovery.

Share

COinS