Data Driven AI Framework for Improving Healthcare Outcomes using Natural Language Processing and Machine Learning Techniques

Abstract

The adoption of Artificial Intelligence (AI) in healthcare has created new opportunities for improving medical diagnosis, patient support, and treatment planning. This paper proposes a data-centric AI framework that integrates Natural Language Processing (NLP), Machine Learning, and Generative AI to deliver intelligent healthcare assistance. The framework analyzes patient-reported symptoms using TF-IDF feature extraction and a Logistic Regression classifier to identify potential diseases. Furthermore, a Generative AI component generates comprehensive medical information, including disease descriptions, risk assessments, and preventive recommendations. The proposed solution is implemented as a Flask-based web application featuring secure user authentication and conversation history management. Experimental evaluation demonstrates that the framework can accurately classify diseases from textual symptom descriptions while providing informative healthcare guidance. The system improves access to preliminary medical support, minimizes dependence on manual assessment, and enhances healthcare decision-making. Overall, the proposed framework contributes toward the development of intelligent, scalable, and user-friendly healthcare technologies.