Detecting Autism Spectrum Disorder Using Machine Learning

This project explores the use of eye-tracking technology and machine learning to detect Autism Spectrum Disorder (ASD) in children, a condition that affects social communication and behavior. Early diagnosis of ASD is critical as it enables timely interventions, improving the quality of life for affected children and their families. The developed system analyzes eye movement patterns using Python-based algorithms and machine-learning models, with a focus on vector machine classifiers and computer vision techniques like the Viola-Jones algorithm. The primary outcome of this project is an efficient, automated diagnostic tool capable of detecting ASD with an accuracy of 89%. The system evaluates subtle variations in eye movements that are challenging to observe with the naked eye. These variations are critical markers for ASD diagnosis, as children with ASD often exhibit distinct eye gaze patterns compared to neurotypical peers. The project also demonstrates the ability to classify images of children into ASD and non-ASD categories based on their eye tracking data. This classification not only aids in diagnosis but also sets a foundation for further research into ASD-related behavioral patterns. By automating the diagnostic process, the tool minimizes subjectivity and human error, offering a reliable alternative to traditional assessment methods. The outcomes of this project include a significant reduction in diagnostic time, improved accuracy, and a scalable system that can be deployed in clinical and research settings. This innovative approach has the potential to revolutionize ASD diagnosis, making early detection more accessible and cost-effective, ultimately contributing to better developmental outcomes for children.