Development and validation of MRI-based model for the preoperative prediction of macrotrabecular hepatocellular carcinoma subtype

Abstract

Key points Clinicoradiologic model has a good performance in diagnosing MTHCC. In validation set, models 2 and 3 did not significantly improve predictive performance compared to clinicoradiologic model. Clinicoradiologic model yielded higher clinical benefit at MTHCC threshold probability of 10%. Clinicoradiologic model is as beneficial as model 3, yet simpler and easier.

Keywords

• Macrotrabecular hepatocellular carcinoma
• Magnetic resonance imaging
• Nomograms
• Decision-making