End-to-End Cascaded U-Nets with a Localization Network for Kidney Tumor Segmentation

TL;DR

This paper introduces TuNet, a cascaded 3D U-Net approach with a localization network for improved kidney and tumor segmentation in medical imaging, demonstrating competitive results on the 2019 Kidney Tumor Segmentation Challenge dataset.

Contribution

It proposes a novel cascaded 3D U-Net architecture with a localization network for sequential kidney and tumor segmentation, advancing the state-of-the-art in this challenging task.

Findings

Achieved a 0.902 Dice score for kidney segmentation.

Achieved a 0.408 Dice score for tumor segmentation.

Validated on 210 patient scans with five-fold cross-validation.

Abstract

Kidney tumor segmentation emerges as a new frontier of computer vision in medical imaging. This is partly due to its challenging manual annotation and great medical impact. Within the scope of the Kidney Tumor Segmentation Challenge 2019, that is aiming at combined kidney and tumor segmentation, this work proposes a novel combination of 3D U-Nets---collectively denoted TuNet---utilizing the resulting kidney masks for the consecutive tumor segmentation. The proposed method achieves a S{\o}rensen-Dice coefficient score of 0.902 for the kidney, and 0.408 for the tumor segmentation, computed from a five-fold cross-validation on the 210 patients available in the data.