The application of artificial intelligence in the intersection of metabolic dysfunction-associated steatotic liver disease and cardiovascular diseases

There is high comorbidity and complex pathological mechanisms between metabolic dysfunction−associated steatotic liver disease (MASLD) and cardiovascular disease (CVD), and the accuracy of traditional risk assessment tools is insufficient. The paper highlights that artificial intelligence (AI) including machine learning and deep learning capable of integrating clinical, imaging, and multi-omics data to enhance the precision of diagnosing MASLD and staging liver fibrosis, the related model has AUC greater than 0.85, and moreover, AI can also accurately predict CVD risk of patients with MASLD, which related model has AUC greater than 0.8 and whose performance is better than traditional scoring systems. In the medical field, deep learning facilitates the quantification of liver fat, along with the evaluation of coronary plaque and screening for lesions across different organs. Multimodal AI has the potential to reveal novel mechanisms and biomarkers of diseases. In addition to these challenges which include data quality and model generalization, the paper also points to future directions such as federated learning. AI offers a fresh perspective on assessing risks, understanding mechanisms, and implementing clinical interventions for MASLD-CVD.