Artificial Intelligence–driven Approaches For Liver Disease Diagnosis And Prediction: A Comprehensive Survey

Abstract

Liver Diseases, Including Liver Cirrhosis And Cholangiocarcinoma, Are The Composites Formed In The Binding Process Of Buried Solidic aggregates Into Light And Porous Preforms. Epidemiology Analyses Suggest Increasing Disease Burden And long-term Trends In Progression Internationally. Early Diagnosis Is Essential, But The Current Diagnostic Methods, Such As Biopsy And Imaging Interpretation, Are Invasive, Expensive, And Subjective To Inter-observer Variability. Recent Developments In Artificial Intelligence (AI) And Specifically The Subfield Of Machine Learning (ML), Including Deep Learning (DL), Have Led To Automated Data-driven Systems For Better Diagnostic Accuracy And Prognosis Prediction Of Diseases. The Genomic Modeling Based On LightGBM Has A Good Performance To Diagnose cholangiocarcinoma, And Ensemble Learning Methods Can Stabilize Prediction For Structured Clinical Datasets. Very Recently, Methods Based On Deep Learning Networks That Utilize EfficientNet-B7 And Dual Attention Mechanisms Have Attained High Performance For Fibrosis Staging. Additionally, Explainable AI (XAI) Frameworks Like XAIHO Advance Interpretability And Clinical Trust, Which Is Consistent With General Healthcare AI Transparency Principles. However, The Data Imbalance, Overfitting, and Lack Of Interpretability Problems Remain. In This Review, A Comprehensive Collection And Synthesis Of The State-of-the-art AI-based Approaches Is Summarized, And Methodological Trends Are reviewed Alongside Future Research Topics Such As Multimodal Fusion Or The Development Of Explainable, Clinically Validated AI Systems.

Keywords