Hybrid Intelligent Model for Prediction of Autism Spectrum Disorder in Children

Autism Spectrum Disorder (ASD) presents a significant challenge in early diagnosis and intervention due to its complex and varied symptomatology. ASD poses significant challenges in early detection and intervention due to its multifaceted nature. This study presents a Hybrid Intelligent Model designed to predict ASD in pediatric cases, leveraging adaptive neuro-fuzzy systems. The model integrates artificial neural network capabilities with fuzzy logic, offering a comprehensive approach to ASD prediction. A diverse dataset comprising behavioral observations, developmental milestones, and clinical assessments is utilized to identify key features relevant to ASD diagnosis. These features include eye contact, gesture use, language skills, sensitivity to pain, communication abilities, and social interaction. Through fuzzy logic-based soft computing techniques, the model achieves enhanced accuracy in predicting ASD and assessing its severity in children. Sensitivity analysis highlights the significant contributions of input variables to ASD prediction, with sensitivity to pain, eye contact level, and social interaction emerging as crucial factors. Comparative analysis with the Back Propagation Algorithm underscores the superiority of the proposed Hybrid Algorithm in error minimization across various phases of model training and evaluation. The findings underscore the potential of adaptive neuro-fuzzy systems in facilitating early ASD diagnosis, enabling timely intervention and support for affected children and their families. This research contributes to advancing the understanding and management of ASD, offering valuable insights for clinical practice and research in pediatric neurodevelopmental disorders.