Key points are not available for this paper at this time.
Cirrhosis is frequently associated with portal hypertension (PH), which can result in severe complications, including varices, ascites, and hepatic encephalopathy. The gold standard for diagnosing PH is the hepatic venous pressure gradient; however, its invasive nature necessitates the exploration of noninvasive diagnostic alternatives. Imaging techniques such as ultrasound, computed tomography, and magnetic resonance imaging are frequently employed but encounter challenges in the early detection of clinically significant portal hypertension (CSPH). Recent advances in artificial intelligence (AI), particularly machine and deep learning, have provided promising solutions. AI can analyze complex medical images and enhance diagnostic accuracy by identifying early indicators of PH. Techniques such as radiomics and vascularomics have demonstrated high efficacy in predicting CSPH, thereby improving noninvasive assessment. AI integration with multimodal data may yield more comprehensive, precise, and noninvasive diagnostic tools, facilitating early detection and enhancing the treatment of cirrhosis-related complications.
Du et al. (Thu,) studied this question.