🤖 AI Summary
To address the high cost of voxel-level annotations in 3D medical image segmentation, this paper proposes TraCoCo, a semi-supervised framework. Unlike mainstream consistency-based methods relying on spatially invariant perturbations (e.g., noise addition or cropping), TraCoCo introduces **spatial translation perturbations**, which deliberately disrupt background contextual consistency and thereby compel the model to focus on foreground structures. Furthermore, it designs a **translation-consistent co-training mechanism** and a **confidence-region cross-entropy (CRC) loss**, jointly improving pseudo-label quality and training stability. Evaluated on four major benchmarks—Left Atrium (LA), Pancreas, BraTS19, and ACDC—TraCoCo achieves state-of-the-art performance. The source code and pre-trained models are publicly available.
📝 Abstract
3D medical image segmentation methods have been successful, but their dependence on large amounts of voxel-level annotated data is a disadvantage that needs to be addressed given the high cost to obtain such annotation. Semi-supervised learning (SSL) solves this issue by training models with a large unlabelled and a small labelled dataset. The most successful SSL approaches are based on consistency learning that minimises the distance between model responses obtained from perturbed views of the unlabelled data. These perturbations usually keep the spatial input context between views fairly consistent, which may cause the model to learn segmentation patterns from the spatial input contexts instead of the foreground objects. In this paper, we introduce the Translation Consistent Co-training (TraCoCo) which is a consistency learning SSL method that perturbs the input data views by varying their spatial input context, allowing the model to learn segmentation patterns from foreground objects. Furthermore, we propose a new Confident Regional Cross entropy (CRC) loss, which improves training convergence and keeps the robustness to co-training pseudo-labelling mistakes. Our method yields state-of-the-art (SOTA) results for several 3D data benchmarks, such as the Left Atrium (LA), Pancreas-CT (Pancreas), and Brain Tumor Segmentation (BraTS19). Our method also attains best results on a 2D-slice benchmark, namely the Automated Cardiac Diagnosis Challenge (ACDC), further demonstrating its effectiveness. Our code, training logs and checkpoints are available at https://github.com/yyliu01/TraCoCo.
