AI-Based Early Detection of Neurological and Mental Disorders using Multimodal Deep Learning

Abstract

Early diagnosis of neurological and mental health disorders is a pressing issue in the current healthcare system, especially in underdeveloped areas with limited access to advanced diagnostic tools and healthcare professionals. Conditions like Parkinson's, depression, and stroke often show early symptoms that are challenging to detect using traditional medical approaches. In this research, we introduce a new non-invasive and affordable multimodal deep learning model that uses both speech and facial micro-expressions to predict disease in a timely and efficient manner. Our proposed system uses Convolutional Neural Networks (CNN) to extract spatial features from facial images and Recurrent Neural Networks (RNN/LSTM) to model temporal speech features. A fusion strategy is used to merge multimodal features, improving the model's accuracy and reliability. This approach aims to identify individuals at various risk levels, facilitating early detection and ongoing health monitoring. Moreover, this study highlights the practicality of the system in real-world settings, especially in underserved rural areas, by minimising the need for costly diagnostic tests. We also discuss the issues of privacy, dataset collection, and generalisation. The method has the potential to revolutionize preventive health by facilitating scalable, affordable and smart health diagnostic and monitoring systems.