Key Points The use of a fixed HVPG value (2.5 mmHg) for healthy controls may introduce systematic bias, particularly near the 6 mmHg clinical threshold. Excluding patients with portal vein thrombosis—present in 15%–25% of cirrhosis—limits the real‐world applicability of current radiomics models. Integrating non‐invasive parameters endorsed by Baveno VII (e.g., liver/spleen stiffness) and CT‐derived body composition metrics could improve risk stratification.
Li et al. (Tue,) studied this question.