Skip to main content

Table 3 Experimental results for liver and tumor segmentation with filter size reduction and application of data augmentation

From: Modified U-Net for liver cancer segmentation from computed tomography images with a new class balancing method

 

Liver

(In DSC)

Tumor from liver

(In DSC)

Tumor from abdominal CT image

(In DSC)

UNet (with original filter size)

0.9529

0.7384

0.6743

Modified UNet without data augmentation

0.9027

0.0992

0.0287

Modified UNet with data augmentation.

0.9612

0.74

0.63