SafeMeds - Drug Interaction Predictor

Drug–drug interactions are a major problem in healthcare, especially when patients take multiple medicines at the same time. Traditional methods for detecting these interactions are slow, depend on manual checking, and may miss complex or new risks. This project, SafeMeds, is an AI-based system developed to quickly and accurately identify harmful drug combinations. It uses a web interface built with HTML, CSS, and JavaScript, and a backend developed using Python Flask and SQLite. Machine Learning and Deep Learning models like CatBoost and LSTM are used to analyze drug data and predict interaction risks. Users can enter drug names along with details such as dosage, age, gender, and existing diseases. The system then predicts whether the interaction risk is High, Moderate, or Low, along with a confidence score. It also uses SHAP to explain the results, making the system transparent and easy to understand. SafeMeds helps improve patient safety, saves time, and supports better medical decision-making.