ConPro: Contrast Projection Pretraining for Label-Efficient Vessel Segmentation in DSA Sequences

2026-09-24

AI Summary

This study addresses the high annotation costs of DSA vessel segmentation and the underutilization of dynamic contrast signals in existing methods by proposing ConPro, a self-supervised pre-training scheme. ConPro pioneers the construction of contrast projections—defined as the normalized decline relative to the pixel-wise temporal median—as the learning objective. It operates in a plug-and-play manner without modifying the segmentation architecture, and we demonstrate that its performance gains stem from this target design rather than generic reconstruction or pseudo-labeling. Experiments show that ConPro significantly outperforms training from scratch under low-annotation regimes. Furthermore, when integrated with the UniMatch framework, it achieves Dice scores of 75.4% and 81.3% on the DIAS and DSCA datasets, respectively.

Abstract

Dense vessel annotation in digital subtraction angiography (DSA) is labor-intensive, yet every unlabeled sequence records how contrast passes through the vessels. Semi-supervised methods take their targets from the current model, and generic self-supervised pretexts reconstruct static appearance, so this signal goes unused. We propose ConPro, a self-supervised pretraining scheme whose target is a contrast projection, the normalized drop of every pixel below its temporal median over the sequence. On DIAS and DSCA, with 10%, 20% and 50% of the training cases labeled, ConPro improves on training from scratch at every label fraction and is the best of the compared methods on DSCA at 20% and 50% labels. Controlled comparisons show that the gain comes from the target. A temporal-median target with the same input, loss and budget stays at scratch level, and using the projection directly instead of learning it, as an input channel or a pseudo-label, helps little or hurts. ConPro provides pretrained weights without changing the segmentation architecture, so it combines with semi-supervised training, and UniMatch, the strongest baseline, gains 0.5 to 2.0 Dice and 0.9 to 2.3 clDice at every label fraction when started from ConPro weights, reaching 75.4 Dice on DIAS and 81.3 on DSCA.

Problem

• Vessel Segmentation
• Digital Subtraction Angiography
• Label-Efficient Learning
• Self-Supervised Pretraining

Innovation

• Self-supervised pretraining
• Contrast projection
• Vessel segmentation
• Digital subtraction angiography
• Label-efficient learning