Explainable artificial intelligence in healthcare diagnostics: Enhancing transparency and trust

Abstract

Healthcare is using artificial intelligence (AI) more and more to help with diagnosis and disease prediction. Nevertheless, a lot of AI models are not interpretable, which restricts their usefulness in therapeutic settings. By making model conclusions comprehensible to people, Explainable Artificial Intelligence (XAI) overcomes this constraint.This study presents a survey-based analysis of important XAI techniques, such as SHAP, LIME, and Grad-CAM, and their use in medical diagnostics. The study emphasizas how explainablity enhances decision-making,transparency and trust.Important issues including computing,complexity,bias, and lack of uniformity are also covered.