Pixel-level Counterfactual Contrastive Learning for Medical Image Segmentation

2026-03-17

AI Summary

This work addresses the challenge of medical image segmentation under limited annotated data by extending counterfactual contrastive learning to pixel-level tasks. The authors propose dense contrastive frameworks with dual-view (DVD-CL) and multi-view (MVD-CL) formulations, along with supervised variants that incorporate silver-standard labels to enhance robustness. In fully unsupervised settings, the proposed methods outperform existing dense contrastive approaches; when fine-tuned with silver-standard labels, they achieve approximately 94% Dice Similarity Coefficient (DSC) on complex datasets, significantly surpassing models trained directly on silver-standard annotations. Additionally, the study introduces CHRO-map, a novel visualization algorithm that facilitates interpretability analysis of learned representations.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Image segmentation relies on large annotated datasets, which are expensive and slow to produce. Silver-standard (AI-generated) labels are easier to obtain, but they risk introducing bias. Self-supervised learning, needing only images, has become key for pre-training. Recent work combining contrastive learning with counterfactual generation improves representation learning for classification but does not readily extend to pixel-level tasks. We propose a pipeline combining counterfactual generation with dense contrastive learning via Dual-View (DVD-CL) and Multi-View (MVD-CL) methods, along with supervised variants that utilize available silver-standard annotations. A new visualisation algorithm, the Color-coded High Resolution Overlay map (CHRO-map) is also introduced. Experiments show annotation-free DVD-CL outperforms other dense contrastive learning methods, while supervised variants using silver-standard labels outperform training on the silver-standard labeled data directly, achieving $\sim$94% DSC on challenging data. These results highlight that pixel-level contrastive learning, enhanced by counterfactuals and silver-standard annotations, improves robustness to acquisition and pathological variations.

Problem

• medical image segmentation
• silver-standard labels
• pixel-level contrastive learning
• counterfactual generation
• annotation scarcity

Innovation

• counterfactual generation
• dense contrastive learning
• medical image segmentation
• silver-standard labels
• CHRO-map