Exploring Cycle Consistency Learning in Interactive Volume Segmentation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Qin, Zheng, Meng, Planche, Benjamin, Gao, Zhongpai, Chen, Terrence, Niethammer, Marc, Wu, Ziyan
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909132860162048
author Liu, Qin
Zheng, Meng
Planche, Benjamin
Gao, Zhongpai
Chen, Terrence
Niethammer, Marc
Wu, Ziyan
author_facet Liu, Qin
Zheng, Meng
Planche, Benjamin
Gao, Zhongpai
Chen, Terrence
Niethammer, Marc
Wu, Ziyan
contents Automatic medical volume segmentation often lacks clinical accuracy, necessitating further refinement. In this work, we interactively approach medical volume segmentation via two decoupled modules: interaction-to-segmentation and segmentation propagation. Given a medical volume, a user first segments a slice (or several slices) via the interaction module and then propagates the segmentation(s) to the remaining slices. The user may repeat this process multiple times until a sufficiently high volume segmentation quality is achieved. However, due to the lack of human correction during propagation, segmentation errors are prone to accumulate in the intermediate slices and may lead to sub-optimal performance. To alleviate this issue, we propose a simple yet effective cycle consistency loss that regularizes an intermediate segmentation by referencing the accurate segmentation in the starting slice. To this end, we introduce a backward segmentation path that propagates the intermediate segmentation back to the starting slice using the same propagation network. With cycle consistency training, the propagation network is better regularized than in standard forward-only training approaches. Evaluation results on challenging AbdomenCT-1K and OAI-ZIB datasets demonstrate the effectiveness of our method.
format Preprint
id arxiv_https___arxiv_org_abs_2303_06493
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Exploring Cycle Consistency Learning in Interactive Volume Segmentation
Liu, Qin
Zheng, Meng
Planche, Benjamin
Gao, Zhongpai
Chen, Terrence
Niethammer, Marc
Wu, Ziyan
Computer Vision and Pattern Recognition
Automatic medical volume segmentation often lacks clinical accuracy, necessitating further refinement. In this work, we interactively approach medical volume segmentation via two decoupled modules: interaction-to-segmentation and segmentation propagation. Given a medical volume, a user first segments a slice (or several slices) via the interaction module and then propagates the segmentation(s) to the remaining slices. The user may repeat this process multiple times until a sufficiently high volume segmentation quality is achieved. However, due to the lack of human correction during propagation, segmentation errors are prone to accumulate in the intermediate slices and may lead to sub-optimal performance. To alleviate this issue, we propose a simple yet effective cycle consistency loss that regularizes an intermediate segmentation by referencing the accurate segmentation in the starting slice. To this end, we introduce a backward segmentation path that propagates the intermediate segmentation back to the starting slice using the same propagation network. With cycle consistency training, the propagation network is better regularized than in standard forward-only training approaches. Evaluation results on challenging AbdomenCT-1K and OAI-ZIB datasets demonstrate the effectiveness of our method.
title Exploring Cycle Consistency Learning in Interactive Volume Segmentation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2303.06493