Explainable AI and Machine Learning for Early Diagnosis of Liver Cirrhosis: A Comprehensive Review

Liver cirrhosis is a chronic, progressive, and po- tentially fatal disease resulting from chronic liver damage. The detection in early phase asymptomatic difficult, which making timely diagnosis difficult and caused the sever complication or mortality. Due to Block-box nature of the machine learning model poses a challenge for clinical practitioners to lies these ML models. Liver disease prediction is major clinical issue, in India approximately 1 million cirrhosis cases reported every year with high mortality and worldwide 58.4 million case reported in 2021 and 1.4 million mortality count. The Global Burden of Disease (GBD) database shows that liver cirrhosis increase 58.21 % from 1990 to 2021 . The use of Explainable AI (XAI) and Machine learning (ML) techniques help in detecting liver cirrhosis disease at early phase, thereby minimizing the severity of the disease. This literature study review recent advances (2015–2025) in XAI and ML models for liver cirrhosis detection, highlighting the role in early diagnosis, risk assessment, and automated medical image analysis. This review also addressing challenges related to model transparency, data heterogeneity, and clinical translation.