A Medical Diagnosis System Based on Explainable Artificial Intelligence: Autism Spectrum Disorder Diagnosis

This paper introduces a new diagnostic system for Autism Spectrum Disorder (ASD) using explainable artificial intelligence (AI). The goal is to develop a reliable and interpretable tool that helps healthcare professionals accurately identify individuals with ASD. The study follows a systematic methodology involving comprehensive data collection, feature engineering, and advanced machine learning algorithms, such as decision trees and support vector machines. By analyzing various patient data, including behavioral observations and medical history, the system identifies important features and patterns associated with ASD. The diagnostic system achieves promising results, with the decision tree model achieving an accuracy of 85% and the support vector machine model achieving 86%. These outcomes demonstrate the potential of the system to accurately identify ASD cases. The clinical relevance and practical implications of the diagnostic system are discussed, emphasizing its ability to enhance the accuracy and efficiency of ASD diagnoses. The paper also identifies limitations and proposes future enhancements, including expanding datasets to cover a wider age range and demographic factors, incorporating additional relevant features such as genetic markers and neuroimaging data, exploring alternative machine learning algorithms, and further advancing explainable AI techniques. Real-world validation and feedback from clinicians and caregivers are crucial for refining the system. Ultimately, this research aims to contribute to timely interventions and improved outcomes for individuals with ASD, providing valuable insights for clinicians, caregivers, and researchers in addressing the challenges of ASD diagnosis.