Title :
Infornation Modeling and Analysis of Event Related Potentials
Author :
Williams, William J. ; Shevrin, Howard ; Marshall, Robert E.
Author_Institution :
Electrical Engineering and Computer Science Department and the Bioengineering Program, University of Michigan
Abstract :
An approach to the analysis of event related potentials (ERP) based on information theory is described. The amount of information in cortical potentials evoked by words briefly presented by tachistoscope is measured by a transinformation method. The inputs to the communication channel are four word categories. One category is composed of pleasant words, another of unpleasant words, both drawn from the Osgood semantic differential research, and two categories are composed of word stimuli related to a psychiatric patient´s complaints. The patients chosen for study exhibit either phobias or pathological grief reactions. The method allows ERP´s to be mapped into transinformation profiles that reveal time intervals during the poststimulus ERP favorable to the transmission of information concerning the word category presented. The transinformation model results for actual and synthetic data are discussed.
Keywords :
Biomedical engineering; Communication channels; Enterprise resource planning; Information analysis; Information theory; Mutual information; Pathology; Psychiatry; Psychology; Voltage; Biomedical Engineering; Evoked Potentials; Humans; Information Theory; Models, Psychological;
Journal_Title :
Biomedical Engineering, IEEE Transactions on
DOI :
10.1109/TBME.1987.325931