Voice and Text-Based Healthcare Chatbot with Real-Time Multilingual Translation

Abstract

The integration of artificial intelligence (AI) in healthcare has paved the way for advanced, accessible, and efficient patient support systems. This paper presents a multilingual, AI-powered medical chatbot designed to predict and respond to infectious disease-related queries using a deep learning approach. The chatbot leverages Long Short-Term Memory (LSTM) networks to process and interpret user inputs in both text and voice formats. Trained on a COVID-19-focused question dataset, the model achieves a high training accuracy of 99%, demonstrating strong predictive capabilities. To overcome language barriers, the chatbot integrates Google Translate API, enabling real-time multilingual communication in English and Telugu, with potential for future language expansion.The system architecture includes modules for user authentication, voice and text interaction, NLP-based processing, multilingual translation, and chat history retrieval. A cloud-based backend ensures scalability, secure data handling, and rapid response generation. The chatbot provides medically relevant responses to users in real-time, enhancing accessibility in remote and multilingual populations. Experimental evaluation confirms the model’s effectiveness across accuracy, response time, and language adaptability. This research contributes to AI-driven healthcare by proposing a scalable and privacy-conscious solution that supports multilingual, multimodal interaction. Future enhancements include expanding language support, integrating with electronic health records (EHRs), and improving contextual understanding through advanced NLP and feedback learning.