AI-Driven System for Symptom-Based Disease Prediction and Review-Informed Medicine Recommendation

Abstract

The use of Artificial Intelligence (AI) in different fields is a major potential for improvement in healthcare, especially in diagnostic processes therapeutic suggestions, and pharmaceutical discovery and development. Overcoming the difficulties in receiving timely medical advice, avoiding inappropriate self-medication, and aiding healthcare workers in lowincome settings requires data-driven and creative solutions. This research presents a Novel AI-powered system aimed at predicting diseases based on user symptoms provided by users dynamically and recommending suitable medications, this is designed to enhance healthcare decision-making. The system starts by taking symptoms from the users to predict the possible disease using a Random forest classification model and label encoder, which has been trained on the Disease dataset, with disease labels managed by the Disease label encoder. After disease prediction, it also filters a comprehensive medicine based on the predicted disease and enriches these recommendations with metadata and sentiment analysis based on user reviews(Excellent, Average, and Poor review percentages). The system then uses a scoring mechanism based on reviews and sentiment to identify and recommend a curated list of the top 4 drugs from the filtered set, Providing the users with prioritized options and critical information like composition and manufacturer, empowering them to consider alternatives based on factors like composition allergies or personal preference. The Analysis showed that the model can predict diseases correctly with 100% accuracy. Further, the system correctly filters and ranks suitable drugs, thus providing an evidence-based model for Disease prediction and correct medicine recommendation. The system has the potential to be a decision support system for healthcare professionals, prevent dangerous self-medication, and act as a helpful preliminary tool in areas with control access to medical professionals, with a final outcome of better patient treatment.

Ähnliche Arbeiten