Robust segmentation of biomedical images remains challenging primarily due to ill-defined object boundaries. Although several loss functions have recently been proposed to enhance object boundary recognition, concerns remain about the susceptibility to unstable annotations resulting from boundary ambiguities. Here, we introduce the Peripheral Prototypical Contrastive (PPC) loss, which leverages prototypical contrastive learning to improve boundary recognition and enable stable training even under noisy annotation conditions. The PPC loss utilizes prototype features obtained from local regions near object boundaries, focusing specifically on ambiguous areas that have been challenging for existing approaches. Extensive experiments across multiple datasets demonstrated that the PPC loss consistently improve segmentation performance even under conditions with noisy labels. A newly proposed strategy, which adaptively determines the local region for prototype calculation, enhanced robustness and outperformed existing methods. Our findings indicate the potential of the PPC loss in facilitating scalable and reliable medical image segmentation, including the development of foundation models with imperfect annotations.
Figures & tables
Figure 1: The overview of the proposed the PPC loss ( LPPC ) to enhance boundary recognition of the DNN.
Chest X-ray (JSRT)
CT (LIDC-IDRI)
MR (BraTS2020)
Lung
Heart
Clavicle
Lung nodule
Brain tumor
Methods
Clean
Noise
Clean
Noise
Clean
Noise
Clean
Noise
Clean
Noise
BCE loss + Dice loss
97.50
93.45
89.77
80.88
86.25
77.91
64.21
59.39
83.35
81.18
(Baseline)
± 0.05
± 0.25
± 1.40
± 1.70
± 1.15
± 1.09
± 0.40
± 0.72
± 0.29
± 0.21
+ Edge detection loss
97.75
93.44
86.02
75.09
86.06
78.92
65.06
59.50
82.77
78.99
± 0.03
± 0.14
± 3.74
± 3.41
± 0.41
± 0.89
± 0.33
± 0.58
± 0.66
± 0.39
Table 1: Comparison of segmentation performance using the Dice coefficient. The best and second-best scores are highlighted in bold and underlined , respectively. Gray cells indicate the degradation compared to the baseline.
Figure 2: Visual comparison of predictions by the DNN trained with noisy labels. The ground truth (GT) is shown in blue. (Left) Results for the clavicles in the JSRT database. (Right) Results for the nodule in LIDC-IDRI.
Figure 3: Visual comparison of the background subregions B . B^Ada was estimated by the DNN trained with noisy labels. (Top) Results for the heart in the JSRT database. (Middle) Results for the clavicles in the JSRT database. (Bottom) Results for the nodule in LIDC-IDRI.
Image segmentation remains fundamentally limited by boundary ambiguity arising from sampling-induced information loss and inherent uncertainty in pixel-wise labeling. Although encoder-decoder architectures such as U-Net achieve strong performance, they often produce overconfident predictions that fail to capture transition-region ambiguity. To address this issue, we propose NoiseUNet, a simple yet effective framework that injects bounded perturbations into skip connections to regularize cross-scale feature fusion. This mechanism enforces robustness to local feature variations and promotes boundary-aware representations. Theoretically, the perturbation induces an implicit fuzzification effect, yielding soft, data-driven memberships without requiring explicit fuzzy modeling. We further introduce ThyR, a real-world thyroid ultrasound dataset with inherently ambiguous boundaries. Experiments demonstrate that NoiseUNet consistently improves both segmentation accuracy and boundary fidelity.
Bisheng Tang, Zhangfeng Ma, Chuchu Zhai +4
School of Information Science and Engineering, Provincial Key Laboratory of Informational Service for Rural Area of Southwestern Hunan, Shaoyang University, Shaoyang 422000, China · Xinshao County People’s Hospital, Shaoyang 422900, China
Split Federated Learning (SplitFed) combines federated and split learning to preserve privacy while reducing client-side computation. However, in medical image segmentation, heterogeneous label quality across clients can significantly degrade performance. We propose SplitFed-CL, a co-learning framework where a global teacher guides local students to detect and refine unreliable annotations. Reliable labels supervise training directly, while unreliable labels are corrected via weighted student--teacher refinement. SplitFed-CL further incorporates consistency regularization for robustness to input perturbations and a trainable weighting module to balance loss terms adaptively. We also introduce a novel difficulty guided strategy to simulate human like boundary centric annotation errors, where the degree of perturbation is governed by shape complexity and the associated annotation difficulty. Experiments on two multiclass segmentation datasets with controlled synthetic noise, together with a binary segmentation dataset containing real-world annotation errors, demonstrate that SplitFed-CL consistently outperforms seven state-of-the-art baselines, yielding improved segmentation quality and robustness.
Zahra Hafezi Kafshgari, Hadi Hadizadeh, Parvaneh Saeedi
School of Engineering Science, Simon Fraser University, Burnaby, BC, Canada
In the era of information explosion, efficiently leveraging large-scale unlabeled data while minimizing the reliance on high-quality pixel-level annotations remains a critical challenge in the field of medical imaging. Semi-supervised learning (SSL) enhances the utilization of unlabeled data by facilitating knowledge transfer, significantly improving the performance of fully supervised models and emerging as a highly promising research direction in medical image analysis. Inspired by the ability of Vision Foundation Models (e.g., SAM-2) to provide rich prior knowledge, we propose SSS (Semi-Supervised SAM-2), a novel approach that leverages SAM-2's robust feature extraction capabilities to uncover latent knowledge in unlabeled medical images, thus effectively enhancing feature support for fully supervised medical image segmentation. Specifically, building upon the single-stream "weak-to-strong" consistency regularization framework, this paper introduces a Discriminative Feature Enhancement (DFE) mechanism to further explore the feature discrepancies introduced by various data augmentation strategies across multiple views. By leveraging feature similarity and dissimilarity across multi-scale augmentation techniques, the method reconstructs and models the features, thereby effectively optimizing the salient regions. Furthermore, a prompt generator is developed that integrates Physical Constraints with a Sliding Window (PCSW) mechanism to generate input prompts for unlabeled data, fulfilling SAM-2's requirement for additional prompts. Extensive experiments demonstrate the superiority of the proposed method for semi-supervised medical image segmentation on two multi-label datasets, i.e., ACDC and BHSD. Notably, SSS achieves an average Dice score of 53.15 on BHSD, surpassing the previous state-of-the-art method by +3.65 Dice. Code will be available at https://github.com/AIGeeksGroup/SSS.