A$^3$DSegNet: Anatomy-aware artifact disentanglement and segmentation network for unpaired segmentation, artifact reduction, and modality translation

TL;DR

This paper introduces A$^3$DSegNet, a novel anatomy-aware network that simultaneously addresses vertebra segmentation, artifact reduction, and modality translation in CBCT images using unpaired CT data, significantly improving segmentation accuracy.

Contribution

The paper presents a new unpaired learning framework that integrates three tasks with knowledge sharing, enabling effective vertebra segmentation in low-quality CBCT images without paired data.

Findings

Achieves an average Dice coefficient of 0.926 for CBCT vertebra segmentation.

Outperforms state-of-the-art methods on large clinical datasets.

Effectively handles unpaired data for multi-task learning in medical imaging.

Abstract

Spinal surgery planning necessitates automatic segmentation of vertebrae in cone-beam computed tomography (CBCT), an intraoperative imaging modality that is widely used in intervention. However, CBCT images are of low-quality and artifact-laden due to noise, poor tissue contrast, and the presence of metallic objects, causing vertebra segmentation, even manually, a demanding task. In contrast, there exists a wealth of artifact-free, high quality CT images with vertebra annotations. This motivates us to build a CBCT vertebra segmentation model using unpaired CT images with annotations. To overcome the domain and artifact gaps between CBCT and CT, it is a must to address the three heterogeneous tasks of vertebra segmentation, artifact reduction and modality translation all together. To this, we propose a novel anatomy-aware artifact disentanglement and segmentation network (A$^3$DSegNet) that intensively leverages knowledge sharing of these three tasks to promote learning. Specifically, it takes a random pair of CBCT and CT images as the input and manipulates the synthesis and segmentation via different decoding combinations from the disentangled latent layers. Then, by proposing various forms of consistency among the synthesized images and among segmented vertebrae, the learning is achieved without paired (i.e., anatomically identical) data. Finally, we stack 2D slices together and build 3D networks on top to obtain final 3D segmentation result. Extensive experiments on a large number of clinical CBCT (21,364) and CT (17,089) images show that the proposed A$^3$DSegNet performs significantly better than state-of-the-art competing methods trained independently for each task and, remarkably, it achieves an average Dice coefficient of 0.926 for unpaired 3D CBCT vertebra segmentation.