COMPTUTATIONAL MODELS OF CORTICAL MAP REORGANIZATION

点击翻译按钮获取中文摘要

The broad, long-term objective of this proposed research is to develop, analyze and evaluate computer models of small cortical strokes. In particular, the reorganization of neocortical maps following simulated strokes will be examined. This research will thus complement traditional clinical and animal model approaches to studying cerebrovascular disease. It will shed light on the pathophysiology of post-stroke recovery and on the dynamics of neocortical information processing in general. The methods used in this research are as follows. Computer simulations referred to as neural models will be developed for several neocortical regions known to have topographic and computational maps: primary (SI) and secondary (SII) somatosensory cortex, primary (VI) visual cortex, motor (MI) cortex, and area 7a in parietal cortex. In each case it will be demonstrated that topographic or computational maps resembling those observed empirically emerge spontaneously as a result of synaptic modifications occurring during learning in the neural model. By varying each model region's underlying network structure, activation mechanism, learning method, or other parameters, the role of each of these factors in map formation will be examined. Subsequently, each model will be lesioned ("simulated stroke"). The resultant post-lesion map reorganization will be characterized as lesion size and location are systematically varied. This characterization will make testable predictions that can be investigated in future clinical or laboratory studies. This work will be done by a multidisciplinary research group with expertise in neurology, neurophysiology and computer science.