The Role of Artificial Intelligence in Diagnosing Drug-Induced Hepatitis: A Systematic Review on Differentiation from Viral Hepatitis

Abstract

Exploring the potential of Artificial Intelligence (AI) in enhancing the diagnosis and differentiation of Drug-Induced Liver Injury (DILI) from viral hepatitis. DILI presents symptoms like viral hepatitis, including elevated liver enzymes, jaundice, and liver dysfunction, complicating diagnosis using traditional methods. Accurate and timely differentiation is critical for improving patient outcomes and addressing global morbidity and mortality associated with liver diseases. A systematic search of PubMed and Google Scholar identified 933 studies on AI applications in differentiating DILI from viral hepatitis. Of these, 55 studies were reviewed, focusing on diverse AI techniques and their diagnostic performance metrics. AI models demonstrated high accuracy in distinguishing DILI from viral hepatitis using clinical data, imaging, and biomarkers. Machine learning algorithms were particularly effective in early diagnosis and prognostic predictions. Advancing AI models with multimodal data integration can enhance diagnosis, identify novel therapeutic targets, and reduce healthcare costs through improved patient outcomes and pharmaceutical efficiencies.