A Novel Hybrid Convolutional Neural Network for Accurate Organ Segmentation in 3D Head and Neck CT Images

TL;DR

This paper introduces OrganNet2.5D, a hybrid 2D-3D CNN that improves organ segmentation accuracy in 3D head and neck CT images by addressing resolution disparities and class imbalance issues.

Contribution

The paper presents a novel hybrid CNN architecture combining 2D and 3D convolutions with fewer downsampling layers and hybrid dilated convolutions for better segmentation of organs-at-risk.

Findings

Achieves promising performance on MICCAI 2015 dataset

Outperforms several state-of-the-art methods

Effectively handles resolution differences and class imbalance

Abstract

Radiation therapy (RT) is widely employed in the clinic for the treatment of head and neck (HaN) cancers. An essential step of RT planning is the accurate segmentation of various organs-at-risks (OARs) in HaN CT images. Nevertheless, segmenting OARs manually is time-consuming, tedious, and error-prone considering that typical HaN CT images contain tens to hundreds of slices. Automated segmentation algorithms are urgently required. Recently, convolutional neural networks (CNNs) have been extensively investigated on this task. Particularly, 3D CNNs are frequently adopted to process 3D HaN CT images. There are two issues with na\"ive 3D CNNs. First, the depth resolution of 3D CT images is usually several times lower than the in-plane resolution. Direct employment of 3D CNNs without distinguishing this difference can lead to the extraction of distorted image features and influence the final segmentation performance. Second, a severe class imbalance problem exists, and large organs can be orders of times larger than small organs. It is difficult to simultaneously achieve accurate segmentation for all the organs. To address these issues, we propose a novel hybrid CNN that fuses 2D and 3D convolutions to combat the different spatial resolutions and extract effective edge and semantic features from 3D HaN CT images. To accommodate large and small organs, our final model, named OrganNet2.5D, consists of only two instead of the classic four downsampling operations, and hybrid dilated convolutions are introduced to maintain the respective field. Experiments on the MICCAI 2015 challenge dataset demonstrate that OrganNet2.5D achieves promising performance compared to state-of-the-art methods.