ChatGPT and AI Chatbot for public healthcare applications

Abstract

This project explores the creation of Large Language Model (LLM) chatbot for the public to assess user-described symptoms and offer actionable advice, such as self-care interventions or recommendations to seek medical attention. Additionally, an AI voice bot based on GPT-4o-mini-realtime-preview is an AI psychiatrist specialising in Cognitive Behavioural Therapy (CBT), psychology, and therapeutic guidance and hence provides instant mental health assistance. Developed on a user-friendly interface with Next.js, ReactJS and Tailwind CSS, the application supports both text and voice for seamless and smooth interaction. The application was experimented on with various symptom clusters and mental health situations to identify its accuracy, reliability, and usability. Results indicate that it provides valuable health information and psychological guidance, though it should not be considered a substitute for professional psychiatric or medical advice. This study emphasizes AI's potential to enhance access to extensive health resources and address LLM shortcomings in crucial decision-making processes.