Renal Tumor Segmentation in CT using Cascade U-Net with 2.5D approach

Authors
HE, HE
Shun Leung, Ping
Wang, Lu

Issue Date
2019

Publisher
University of Minnesota Libraries Publishing

Type
Article

Abstract
We propose Cascade U-Net with 2.5D approach to segment kidney and tumor from 3D CT image. We use standard U-Net to generate segmentations per each volume slide (2D image). 4 prediction volumes are generated per different magnification and slice direction. Then, consolidate the volumes to formulate the final prediction volume. Per experiment on the KiTS19 dataset, we get a 12% raise in dice coefficient when compare with single U-Net prediction.

Identifiers
doi: 10.24926/548719.034