Explainable AI-Driven Heart Disease Prediction
R Bhuvaneswari, Pala Mahesh Kumar, Somkuan Kaviya · 2025
The earliest detection is crucial both to prevention and treatment of heart disease, which ranks highly in the list of worldwide health concerns. Data gathering and processing are performed as the initial phase through several health indicators such as blood pressure, cholesterol, and lifestyle factors. All the inputs are processed and computed with the chances of suffering heart disease using a state-of-art model in machine learning. From, this the best performed machine learning model are integrated in the proposed methodology which results in the better accuracy with 83.9%. For the second step, with the help of explainable AI approaches, relevant elements for each prediction of the model are highlighted. Since the performance of the model is assessed in terms of the interpretation and accuracy of its predictions, the third phase comes as the final one, to deliver medical practitioners with an accurate tool that can be put to use for an early diagnosis and helpful insight. The main objective of our approach is to make EAI models in the health care sector more transparent and user-friendly. LIME (Local Interpretable Model-agnostic Explanations) is a popular method in Explainable Artificial Intelligence (EAI) that has been utilized in the proposed work since it interprets complex machine learning models by highlighting the most significant features contributing to a prediction. In the proposed method, LIME provides insights into the features influencing the model's output, such as whether a patient is likely to have heart disease along with the features which contribute to that prediction.