Multi-task Deep Learning Network to Predict Future Macrovascular Invasion in Hepatocellular Carcinoma

We propose a CT-imaging-only prediction framework for future macrovascular invasion in hepatocellular carcinoma. Using graph convolutional networks to capture tumor boundary features combined with global liver structural information, this approach achieves high-precision early prediction without relying on clinical indicators.

Venue: EClinicalMedicine — CAS Q1, JCR Q1 Top, IF: 17.033
Role: Co-first Author

Recommended citation: Haoran Lai*, et al. "Multi-task Deep Learning Network to Predict Future Macrovascular Invasion in Hepatocellular Carcinoma." EClinicalMedicine, 2021.
Download Paper