Abstract

Liver cancer is a pressing global health issue often diagnosed at advanced stages, leading to poor prognosis. Artificial intelligence (AI) has shown promise in improving liver cancer detection from medical images. However, the lack of interpretability in AI models hampers their clinical adop-tion. This research proposes a novel approach that combines U-Net, a deep learning model, with LIME(Local Interpretable Model-agnostic Explanations), Which is an explainable AI technique, to enhance liver cancer detection. The proposed model aims to provide explanations for its predictions,helping physicians comprehend and accept the decisions made by the model.The findings support the development of AI in medical diagnostics by demonstrating the efficacy of the suggested method in identifying liver cancer and offering insightful information about the decision-making process.