Unveiling the Power of Explainable AI: Revolutionizing Healthcare with Transparent Machine Learning and Trustworthy Artificial Intelligence
Abstract
In the field of medicine, ML and deep learning models offer excellent diagnostic results, though their lack of transparency creates obstacles in the area of safety, ethics, and regulations. XAI tackles this transparency issue by explaining how the intelligent systems operate. The paper provides a systematic overview of different XAI techniques that have been used across the fields of CDSS, medical imaging, and EHRs. We perform trade-off evaluations of the applicability of intrinsic XAI approaches versus post-hoc methods (such as SHAP, LIME, and Grad-CAM), check if the methods fit the regulations like EU’s AI Act and FUTURE-AI guidelines, and analyze uncertainty predications through the conformal method. Finally, we outline some critical issues like the difference between correlation and biological causation in XAI.