An Explainable Multimodal AI Framework for Personalized Early Disease Detection Using Medical Images, Clinical Data, and Wearable Sensor Information.

Abstract

Early disease detection plays a vital role in improving patient outcomes and reducing healthcare costs. Recent advances in Artificial Intelligence (AI) have enabled the development of intelligent systems capable of identifying diseases at an early stage using diverse healthcare data. However, existing approaches often rely on a single data source and provide limited interpretability, reducing clinicians' trust in AI-based predictions. This study proposes an explainable multimodal AI framework that integrates medical images, clinical records, and laboratory data to improve the accuracy and reliability of early disease detection. The framework combines deep learning techniques with Explainable Artificial Intelligence (XAI) methods to generate transparent and clinically meaningful predictions. Publicly available healthcare datasets will be used to train and evaluate the proposed model. Performance will be assessed using accuracy, precision, recall, F1-score, and AUC, while explainability will be evaluated using state-of-the-art XAI techniques. The proposed framework aims to support healthcare professionals by providing accurate, interpretable, and timely disease predictions. The findings are expected to contribute to the advancement of trustworthy AI in healthcare and demonstrate the effectiveness of multimodal learning for early disease detection across multiple diseases.

Keywords