fig2

Multi-task neural network enhanced by data augmentation and ROI optimization for prognosis and MVI prediction in HCC using contrast-enhanced CT

Figure 2. Training trajectory and convergence of the proposed model. The final model parameters were fixed at epoch 50 based on optimal validation performance, though the full 80-epoch course is shown to transparently demonstrate the onset of overfitting. (A and B) Prediction accuracy for MVI across the training and validation sets; (C-F) Evolution of the C-index for both OS and RFS throughout the learning process; (G-N) depicts the task-specific loss curves for MVI, OS, and RFS, as well as the total loss, showing the decreasing error rates in both cohorts. MVI: Microvascular invasion; OS: overall survival; RFS: recurrence-free survival; C-index: concordance index.

Hepatoma Research
ISSN 2454-2520 (Online) 2394-5079 (Print)

Portico

All published articles are preserved here permanently:

https://www.portico.org/publishers/oae/

Portico

All published articles are preserved here permanently:

https://www.portico.org/publishers/oae/