Age-Stratified Differences in Morphological Connectivity Patterns in ASD: An sMRI and Machine Learning Approach

Autism spectrum disorder (ASD) is one of the most common neurological disorders, and its early detection is extremely difficult. Researchers use different physiological and medical imaging signals to diagnose ASD based on the severity level and the age of the patient. In this study, morphological features (MF) and morphological connectivity features (MCF) are used to investigate the influence of age on the diagnosis of autism spectrum disorders (ASD). In this work, we have utilized structural magnetic resonance imaging (sMRI) data from ABIDE-I and ABIDE-II databases, divided into 6-11, 11-18, and 6-18 age groups, were pre processed and yielded 592 MF and 10,878 MCF per subject. As a result, the 6-11 age group outperformed the others in both feature types, especially in MCF, with a random forest (RF) classifier achieving 75.8% accuracy, 83.1% F1 score, 86% recall, and 80.4% precision, respectively. Based on this, it can be concluded that an age-specific morphological connectivity approach holds promise for effective diagnosis of autism spectrum disorders.