Stroke is a leading cause of morbidity and mortality, with many patients developing severe disability. We have initiated a series of studies to uncover the neurobiological course of stroke recovery with and without interventions. Neural reorganization is assessed with task dependent BOLD functional magnetic resonance imaging (fMRI).

In this talk, I will focus on two investigations that present particular data analytic problems. First is a longitudinal imaging study of motor stroke recovery, in which patients with strokes are evaluated with fMRI and behavioral measures monthly during the period of maximal recovery. Second is a treatment study of language recovery in which reading therapy administered to a stroke patient was followed with fMRI and language assessment.

Three aspects of these studies present unique problems: (1) Imaging patients is more difficult than imaging normal subjects, since they have more voluntary and involuntary movements, do not tolerate well the scanner, and have abnormal tissue and vessels in the brain; (2) Longitudinal imaging requires anatomical and functional reliability, as well as methods to compare formally the images from different sessions; and (3) Treatment studies are complicated by the issues of reliability, spontaneous recovery, and relating changes in the brain to changes in behavior.

Back to Workshop Schedule