NIRS and NIRI in Motor Tasks
FMRI of Motor Learing
Brain Imaging and Sleep
Neural Net Modeling
I recently completed my PhD dissertation in the department of Cognitive and Linguistic Sciences at Brown University, under the guidance of my major advisor, James. A Anderson. My disseratation abstract appears below, followed by links to a postscript version of the document itself.
This research examined the role of instructional set in the acquisition of serial behavior. In particular, a paradigm was developed to compare two modes of learning a motor sequence: learning practice alone versus learning by instruction plus practice. Three behavioral experiments and one functional magnetic resonance imaging (fMRI) case study are described, all of which address this distinction. The first two behavioral experiments, using a modified serial reaction time task (SRTT), suggest that teaching portions of the structure governing the motor sequence enhance SRTT learning. Furthermore, it was the content of the teaching—and not simply the motivation or attentional correlates—that was important to the learning enhancement. The exact nature of the learning curve changes, and the number of learning systems involved, remained unclear. The third behavioral experiment suggested that both modes of learning depend more on working memory than on attention. That is, learning deficits appeared to grow with added working memory loads, but not so with added attentional distraction. All three behavioral experiments provided evidence that portions of the motor sequence rule structure were learned even when subjects could not report this knowledge. The fMRI experiment sought to examine the neurophysiological correlates of the practice/teaching distinction. The supplementary motor area (SMA) ipsilateral to the motor sequencing task was the primary region exhibiting differences in activation. Ipsilateral SMA displayed elevated but decreasing activation during practice. In contrast, the same region revealed suppressed but increasing activation during practice that followed teaching. The reason for this particular activation pattern is unclear. Nevertheless, the findings suggest that instructions in the Teaching blocks serve to pre-specify different neuronal networks for behavioral performance relative to the Practice blocks.
Postscript version of text (2.0Mb gzipped, 166 pages): strangmandiss.ps.gz
Tarred version of imaging figures (380Kb): strangmandissfig.tar