Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Learning Semi-Supervised Medical Image Segmentation from Spatial Registration

Qianying Liu, Paul Henderson, Xiao Gu, Hang Dai, Fani Deligianni

TL;DR

This paper tackles data scarcity in medical image segmentation by introducing CCT-R, a registration-guided semi-supervised framework. It integrates spatial registration into a contrastive cross-teaching setup via two key modules: Registration Supervision Loss (RSL) and Registration-Enhanced Positive Sampling (REPS). RSL transfers labels through registration transforms between volumes to provide informative pseudo-labels, while REPS augments pixel-level contrastive learning with anatomically corresponding positives across volumes using a memory bank. The approach demonstrates state-of-the-art performance on ACDC and Synapse with very limited labeled data, highlighting the practical impact of leveraging registration information for robust semi-supervised segmentation.

Abstract

Semi-supervised medical image segmentation has shown promise in training models with limited labeled data and abundant unlabeled data. However, state-of-the-art methods ignore a potentially valuable source of unsupervised semantic information -- spatial registration transforms between image volumes. To address this, we propose CCT-R, a contrastive cross-teaching framework incorporating registration information. To leverage the semantic information available in registrations between volume pairs, CCT-R incorporates two proposed modules: Registration Supervision Loss (RSL) and Registration-Enhanced Positive Sampling (REPS). The RSL leverages segmentation knowledge derived from transforms between labeled and unlabeled volume pairs, providing an additional source of pseudo-labels. REPS enhances contrastive learning by identifying anatomically-corresponding positives across volumes using registration transforms. Experimental results on two challenging medical segmentation benchmarks demonstrate the effectiveness and superiority of CCT-R across various semi-supervised settings, with as few as one labeled case. Our code is available at https://github.com/kathyliu579/ContrastiveCross-teachingWithRegistration.

Learning Semi-Supervised Medical Image Segmentation from Spatial Registration

TL;DR

This paper tackles data scarcity in medical image segmentation by introducing CCT-R, a registration-guided semi-supervised framework. It integrates spatial registration into a contrastive cross-teaching setup via two key modules: Registration Supervision Loss (RSL) and Registration-Enhanced Positive Sampling (REPS). RSL transfers labels through registration transforms between volumes to provide informative pseudo-labels, while REPS augments pixel-level contrastive learning with anatomically corresponding positives across volumes using a memory bank. The approach demonstrates state-of-the-art performance on ACDC and Synapse with very limited labeled data, highlighting the practical impact of leveraging registration information for robust semi-supervised segmentation.

Abstract

Semi-supervised medical image segmentation has shown promise in training models with limited labeled data and abundant unlabeled data. However, state-of-the-art methods ignore a potentially valuable source of unsupervised semantic information -- spatial registration transforms between image volumes. To address this, we propose CCT-R, a contrastive cross-teaching framework incorporating registration information. To leverage the semantic information available in registrations between volume pairs, CCT-R incorporates two proposed modules: Registration Supervision Loss (RSL) and Registration-Enhanced Positive Sampling (REPS). The RSL leverages segmentation knowledge derived from transforms between labeled and unlabeled volume pairs, providing an additional source of pseudo-labels. REPS enhances contrastive learning by identifying anatomically-corresponding positives across volumes using registration transforms. Experimental results on two challenging medical segmentation benchmarks demonstrate the effectiveness and superiority of CCT-R across various semi-supervised settings, with as few as one labeled case. Our code is available at https://github.com/kathyliu579/ContrastiveCross-teachingWithRegistration.
Paper Structure (29 sections, 13 equations, 4 figures, 7 tables)

This paper contains 29 sections, 13 equations, 4 figures, 7 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Consistency regularization in semi-supervised medical image segmentation.
  4. Medical image registration.
  5. Combining segmentation and registration.
  6. Contrastive learning for segmentation and registration.
  7. Method
  8. Preliminaries
  9. Supervised contrastive learning.
  10. Training and inference.
  11. Learning from spatial registration
  12. Registration supervision loss
  13. Best registration selection strategy.
  14. Registration-enhanced positive sampling
  15. Experiments
  16. ...and 14 more sections

Figures (4)

  • Figure 1: The overall architecture of our framework for semi-supervised medical image segmentation.
  • Figure 2: Supervised contrastive learning guided by labels vs. registration: In the semi-supervised setting, for unlabeled data, the supervised contrastive loss uses pseudo-label information to select pairs. However, pseudo-labels are unreliable, especially early in training. For example, in the middle panel, the anchor is wrongly labeled as Myo (green), which leads to an incorrect learning signal, due to contrasting with positives correctly labeled as Myo. In contrast, registration finds the anatomically-closest point to the anchor in each 3D volume, without relying on label predictions from models, enabling the contrastive loss to perform correct comparisons between cases.
  • Figure 3: Qualitative results from our CCT-R and baselines. Left: ACDC, trained on 3 labeled cases; right: Synapse, 2 labeled cases
  • Figure 4: DSC of pseudo-labels from two models on unlabeled data during the early training stages, for Synapse (a) 1 labeled case, and (b) 2 labeled cases.