PLESS: Pseudo-Label Enhancement with Spreading Scribbles for Weakly Supervised Segmentation
This work addresses the cost and quality limitations in medical image segmentation for practitioners, though it is incremental as it builds on existing pseudo-label methods.
The paper tackles the problem of noisy and incomplete supervision in weakly supervised segmentation with scribble annotations by proposing PLESS, a pseudo-label enhancement strategy that improves reliability and spatial consistency, leading to consistent accuracy improvements on cardiac MRI datasets.
Weakly supervised learning with scribble annotations uses sparse user-drawn strokes to indicate segmentation labels on a small subset of pixels. This annotation reduces the cost of dense pixel-wise labeling, but suffers inherently from noisy and incomplete supervision. Recent scribble-based approaches in medical image segmentation address this limitation using pseudo-label-based training; however, the quality of the pseudo-labels remains a key performance limit. We propose PLESS, a generic pseudo-label enhancement strategy which improves reliability and spatial consistency. It builds on a hierarchical partitioning of the image into a hierarchy of spatially coherent regions. PLESS propagates scribble information to refine pseudo-labels within semantically coherent regions. The framework is model-agnostic and easily integrates into existing pseudo-label methods. Experiments on two public cardiac MRI datasets (ACDC and MSCMRseg) across four scribble-supervised algorithms show consistent improvements in segmentation accuracy. Code will be made available on GitHub upon acceptance.