Application of adaptive systems methodology to radiotherapy

Abstract

Brain volume differences from 58 children are analyzed to determine the degree of volume loss and the effect on IQ after undergoing radiotherapy for tumors in an effort to identify relationships that might yield knowledge in preventing brain volume loss in future treatments. Analysis of the pre- and post-treatment data is performed first using traditional statistics and then with the assistance of a new kind of artificial adaptive systems called the activation and competition system (ACS) and Auto-contractive Map (Auto-CM). While the result of the statistical study suggests that it is not possible to linearly classify the subjects into subsets of higher and lower IQ, the ACS clearly delineates the dataset into two IQ groups. Further, Auto-CM allows us to establish a semantic connection map among different brain segments which indicates a possible interpretation rule in the observed results. The effect of radiation treatment on the nine brain segments is addressed and future research directions are introduced.