XSSR: Cross-Domain Self-Supervised Representative Selection for Efficient Annotation in Medical Image Segmentation

2026-06-02

AI Summary

This work addresses the challenges of high annotation costs and domain shifts—arising from differences in imaging devices or patient populations—in cross-domain medical image segmentation. The authors propose a three-stage, low-budget sample selection framework: first, a shared embedding space is constructed using a Masked Autoencoder (MAE) trained on unlabeled source data; second, a greedy algorithm selects representative target samples based on density, novelty, and diversity criteria, with the trade-off parameter α automatically calibrated; third, a U-Net is trained exclusively on the selected samples. Requiring no manual hyperparameter tuning, the method achieves 99.3% of fully supervised performance on chest X-ray segmentation with only 5% annotation budget (22 labeled cases), surpasses random selection by 2.5 Dice points on prostate MRI, and consistently outperforms the CoreSet baseline by 0.4–1.2 Dice points across all datasets.

Abstract

Acquiring labeled medical image data is resource-intensive and a challenge further exacerbated in cross-domain scenarios where source and target datasets differ in imaging equipment, population, or clinical site. This study introduces XSSR (Cross-Domain Self-Supervised Representative Selection), a framework designed to minimize annotation effort in the target domain while maintaining robust segmentation performance. XSSR comprises three stages: first, a Masked Autoencoder (MAE) is trained on unlabeled source data to establish a shared embedding space without requiring target labels; second, a greedy selection algorithm scores unlabeled target samples based on a composite density, novelty, and diversity criterion; and third, a U-Net segmentation model is trained exclusively on the selected subset. The novelty-diversity trade-off parameter, alpha, is automatically calibrated by minimizing embedding-space coverage, eliminating manual tuning. We evaluate XSSR on three public benchmarks: Chest X-ray, RIGA+ retinal fundus imaging, and multi-site Prostate MRI, each under a fixed 5% annotation budget. XSSR achieves 99.3% of full-data performance on Chest X-ray using only 22 labeled samples, surpasses random selection by up to 2.5 Dice points on Prostate MRI, and consistently outperforms the CoreSet baseline by 0.4 to 1.2 Dice points across all datasets. Ablation studies indicate that diversity is the most influential scoring component, and per-site analysis shows that performance correlates with scanner similarity to the source domain.

Problem

• cross-domain
• medical image segmentation
• annotation efficiency
• representative selection
• limited labeling

Innovation

• cross-domain
• self-supervised learning
• representative selection
• medical image segmentation
• annotation efficiency