The Application of Computer Techniques to Clinical Research in Convulsive Disorders

Publisher Website
Google Scholar
Abstract
In order to explore the usefulness of general purpose computers for clinical neurological research, a group of 57 epileptic patients was coded in detail in regard to neurologic, psychiatric, psychologic and electroencephalographic findings. 1152 different variables were recorded from each patient and punched on cards. The frequency distribution of each variable was then obtained, and 155 were subsequently utilized for intercorrelations. 1089 statistically significant correlations were obtained. 88 variables with the highest correlation coefficients were then subjected to factor analysis and 11 factors were extracted. In view of the vast amount of information obtained, this paper presents at this time only some examples of the results. These deal with the correlations obtained with the variable "history of grand mal status epilepticus" and as a second example with the variable "atypical spike wave activity in resting EEG." Factor I dealing with the relationships of brain damage to severe seizure disorders and Factor in dealing with relationships of psychotic symptoms in the epileptic patient are shown as results of factor analysis. The study confirmed the usefulness of automatic data processing techniques for clinical research and opens the way for further investigations involving a greater number of patients.

This publication has 10 references indexed in Scilit: