Skip to main content

Table 2 Function strategy in the HFS-Net

From: A hierarchical fusion strategy of deep learning networks for detection and segmentation of hepatocellular carcinoma from computed tomography images

 

Phase

Neural network

Loss function

fliver

Portal-venous

2D DenseU-Net

cross entropy + dice loss

fsize

Dynamic

2D DenseU-Net

focal loss + dice loss

flarge

Portal-venous

2D U-Net

cross entropy

fsmall

Dynamic

2D DenseU-Net

focal loss + dice loss

f3D

Portal-venous

3D U-Net

focal loss + dice loss

  1. Dynamic CT images include non-contrast, arterial, and portal-venous phases
Back to article page

Cancer Imaging

ISSN: 1470-7330

Contact us