Wearable Health Monitoring System for Detecting Physiological Changes in Individual with Autism Spectrum Disorder Along with Anxiety Detection

Abstract

This project introduces a Comprehensive Autism and Health Assessment System created on the basis of the deep learning and machine learning techniques. The system is created to act as a pre-screening instrument that will combine three significant examinations into one web-based application created with Streamlit. The former module does the autism detection with the help of a YOLOv8 object detector model, which evaluates the uploaded image and shows the outcome of the detection and the confidence level. In case of autism identification, the system passes to the second module, which assesses respiratory health conditions through a Support Vector Machine model upon the input of vital signs and symptoms as typed by the user. The condition is categorized as either normal, mild, severe or chronic in the output. The third module is based on anxiety measurement and predicts Separation Anxiety Disorder and Generalized Anxiety Disorder with the help of machine learning pipelines that rely on the inputs of behavioral and emotional symptoms. The system adheres to a step-by-step working process and introduces a final combined report that would show all the outcomes and fundamental recommendations. The application is not to be used professionally and is only to be used as a learning and research tool. The system illustrates that it is possible to have several artificial intelligence models integrated into one viable healthcare screening platform.