CYBERNETIC APPROACH IN IDENTIFICATION OF BRAIN PATTERN VARIATIONS IN AUTISM SPECTRUM DISORDER

Abstract

Poincar section is a tool used in analysis and even control of non-linear systems like chaotic and uncertain systems. Although it has been presented long ago, yet this approach is artistic and heuristic. Poincar Section is destitute of any definite methodologies and problems including indefinite structure and model parameters that can be generally attributed to this approach; machine learning based on Poincar section is impossible. In this article, first of all, signal modeling steps using Poincar is explained, then considering the occurred events, the concept of information and relativism applying Poincar section and information approach, we will diagnose the brain pattern variations in Autistic cases. The reason we have taken Autism into consideration is because we believe its origin is information, in other words the big problem in Autism disorder is software kind, which can lead to hardware kind over time. In this research a new kind of representation, namely Extended Complementary Plot, in which the main characteristic is special attention to signal phase as embedded information in the signal and ineffectiveness of energy, is introduced. All the introduced state-of-art concepts on Electroencephalography are implemented on Autistic children. Recording the EEG signal in Autistic children has always been a challenge for the specialists. Implementations of the article have been carried out on over 120 cases including 60 Autistic children and 60 normal ones ranging from 3 to 10 years old, in three different states; asleep, open eyes and a new record based on brain dynamics which has been suggested from the authors and does not have the other records problems for Autistic kids. Prodigious results accomplished, suggests the common dynamic presence in Autism disorder which is entirely different from normal dynamics, and this is only due to the potency of the applied information tool; Poincar section, and cybernetic modeling in this research. We hope that the empirical results of this research to be a strong and effective step towards quantification of Autism disorder and conversion of diagnosis process from Clinical to Para clinical, and even early Autism diagnosis.