AI-Chatbot for Disease Prediction

Abstract

This AI-based medical Chabot, particularly in disease prediction, highlighting the limitations of earlier Chabot’s that rely on Long Short-Term Memory (LSTM) networks, such as narrow disease coverage and reduced accuracy. Recurrent neural networks (RNNs) are proposed as a promising alternative due to their enhanced capabilities in processing complex patient queries and providing accurate health recommendations. A systematic review emphasizes the need for medical Chabot’s to integrate extensive disease databases and advanced natural language processing (NLP) techniques to improve interaction quality. Despite NLP’s potential, current systems face challenges like language asymmetry, leading to response inaccuracies. Additionally, emerging trends such as hybrid AI models and modular design approaches are fostering greater adaptability in Chabot development. This survey identifies key challenges, including data dependency, scalability, and the necessity for improved accuracy in real-world applications, suggesting that a shift to RNN-based architectures and broader training data could significantly enhance the practical utility of medical Chabot’s, ultimately improving patient outcomes and healthcare support reliability.