Biomedical image segmentation improved by focusing on uncertain boundaries
Beyond Accuracy: Uncertainty-Guided Boundary Refinement for Reliable Biomedical Image Segmentation
Computer Vision and Pattern RecognitionMachine Learning
Summary
Segmenting biomedical images accurately means not just labeling regions correctly but also precisely outlining important structures like cells. The authors propose a method that first makes a standard prediction and then refines only the uncertain boundary areas based on multiple uncertainty measures. This approach slightly improves accuracy and boundary clarity in blood-smear images. Although it doesn’t improve overall confidence calibration, it provides a transparent way to trust and improve boundary segmentation.
What this means in practice
- •For clinical imaging teams: Enhance segmentation accuracy at cellular boundaries in microscopy images for better downstream morphology analysis.
- •For medical device developers: Build imaging software that selectively refines uncertain segmentation areas, improving product reliability in diagnostic tools.$Commercial implications: Enables devices with improved trustworthiness in boundary detection for clinical diagnostics, a competitive feature for medical imaging products.
Authors
Anima Kujur
Abstract
Accurate biomedical image segmentation requires not only high global overlap but also reliable delineation of clinically meaningful boundaries. In blood-smear microscopy, cytoplasm and nucleus contours provide the structural basis for downstream morphology analysis; however, deep segmentation models may remain uncertain or overconfident near ambiguous boundary regions even when achieving strong Dice scores. This work proposes a Reliability-Aware Boundary Refinement Network (RABR-Net), a two-stage framework for trustworthy image segmentation. A strong UNet++ EfficientNet-B4 base segmenter first produces initial class probabilities and logits. Predictive entropy, test-time augmentation variance, margin uncertainty, probability gradients, and soft boundary cues are then combined into a boundary-aware reliability representation. This representation guides a gated residual refiner that selectively corrects uncertain boundary pixels while preserving confident regions of the base prediction. The framework is evaluated using overlap accuracy, class-wise Dice, Boundary Dice, HD95/ASSD, calibration, risk--coverage analysis, robustness under image perturbations, qualitative correction maps, and paired statistical testing. On the held-out test set, the proposed method improves Dice from 0.9602 to 0.9614, Boundary Dice from 0.3448 to 0.3611, and HD95 from 3.0354 to 2.8274 compared with the cached base prediction. Statistical analysis confirms significant improvements in Dice, Boundary Dice, and HD95. Qualitative results show that the learned gate concentrates around uncertain cytoplasm and nucleus boundaries, and correction maps confirm localized boundary refinement. Although calibration does not automatically improve after refinement, the proposed framework provides an interpretable and reliability-focused strategy for boundary-sensitive biomedical image segmentation.