Attention-Guided Semi-Supervised Segmentation for Low-Contrast Medical Images Using Frequency Domain Priors

Medical image segmentation under limited annotation budgets remains a critical challenge, particularly for anatomical structures exhibiting low-contrast boundaries—a pervasive problem in cardiac MRI, dermoscopy, and ultrasound imaging. Existing semi-supervised methods leverage consistency regularization and pseudo-label strategies but typically treat image preprocessing as a static pipeline step, failing to exploit the rich structural priors encoded in frequency-enhanced representations. In this paper, we propose FDPSeg, a novel Frequency-Domain Prior guided semi-supervised Segmentation framework that integrates Contrast-Limited Adaptive Histogram Equalization (CLAHE)-derived frequency priors directly into the transformer attention mechanism. Our dual-branch encoder processes both the original image and its CLAHE-enhanced counterpart, with a learned channel-attention fusion module that adaptively weights spatial and frequency features per image region. A novel frequency-domain consistency loss enforces structural coherence between teacher and student networks in the Fourier space, providing stronger supervisory signal for unlabeled data than spatial consistency alone. Experiments on the ACDC cardiac MRI dataset, the ISIC 2018 skin lesion dataset, and the BUSI breast ultrasound dataset demonstrate that FDPSeg consistently outperforms state-of-the-art semi-supervised baselines under 5% and 10% labeled data regimes, achieving improvements of up to 2.1% in mean Dice score and 3.8 mm reduction in 95th-percentile Hausdorff Distance (HD95) over the strongest competitor, with particularly pronounced gains on low-contrast boundary regions.