Cylindrical Transform: 3D Semantic Segmentation of Kidneys With Limited Annotated Images

TL;DR

This paper introduces a cylindrical transform technique for augmenting limited 3D kidney CT images, enabling the use of classification DCNNs for improved semantic segmentation by capturing spatial and sequential dependencies.

Contribution

The novel cylindrical transform method allows effective 3D data augmentation, facilitating the use of classification DCNNs over traditional FCNs for kidney segmentation with limited annotations.

Findings

Higher segmentation performance with cylindrical transform-augmented data

Classification DCNNs outperform FCNs on limited datasets

Effective utilization of spatial and sequential dependencies

Abstract

In this paper, we propose a novel technique for sampling sequential images using a cylindrical transform in a cylindrical coordinate system for kidney semantic segmentation in abdominal computed tomography (CT). The images generated from a cylindrical transform augment a limited annotated set of images in three dimensions. This approach enables us to train contemporary classification deep convolutional neural networks (DCNNs) instead of fully convolutional networks (FCNs) for semantic segmentation. Typical semantic segmentation models segment a sequential set of images (e.g. CT or video) by segmenting each image independently. However, the proposed method not only considers the spatial dependency in the x-y plane, but also the spatial sequential dependency along the z-axis. The results show that classification DCNNs, trained on cylindrical transformed images, can achieve a higher segmentation performance value than FCNs using a limited number of annotated images.