An Explainable AI Framework for Transparent Medical Billing: Integrating SHAP and LIME in a Web-Based Decision Support System

Abstract

Understanding how automated predictions are made is crucial for enhancing transparency and user confidence in healthcare billing systems. In order to improve the interpretability of medical billing forecasts, this study presents a web-based Explainable Artificial Intelligence (XAI) platform. Local Interpretable Model-agnostic Explanations (LIME) and SHapley Additive exPlanations (SHAP) are combined in the system to explain the variables that affect model predictions at both the individual-case and general levels. The application, which was developed with Python, FastAPI, and React.js, enables users to upload medical billing data, find possible billing problems, and investigate predicted explanations via an interactive interface. Three machine-learning models-Random Forest, XGBoost, and Logistic Regression-were put into practice and assessed; Random Forest outperformed the other models. It achieved an MSE of 8.75 and a reported $\mathrm{R}^{2}$ value of approximately 0.989. The results suggest that combining predictive models with accessible explanation techniques can support more transparent and understandable AI-assisted medical billing analysis.