AI BASED PATIENT SYMPTOMS COLLECTIONS AND DISEASE PREDICTION

This study introduces an AI-based diagnostic support system leveraging machine learning classifiers (MLCs) and natural language processing (NLP) techniques to analyze electronic health record (EHR) data. By querying EHRs akin to the reasoning process of physicians, the system uncovers associations undetected by traditional statistical methods. Deep learning is employed to extract clinically relevant information from 101.6 million data points encompassing 1,362,559 pediatric patient visits to a major referral center. The model demonstrates high diagnostic accuracy across various organ systems, comparable to experienced pediatricians in identifying common childhood diseases. This research showcases the potential of AI systems to assist physicians in managing large datasets, enhancing diagnostic evaluations, and providing decision support in cases of diagnostic uncertainty or complexity, with implications for healthcare providers worldwide. This paper introduces an AI-based system for collecting patient symptoms and predicting diseases.