HELPNet: Hierarchical Perturbations Consistency and Entropy-guided Ensemble for Scribble Supervised Medical Image Segmentation
This work addresses the high cost of full annotations in medical imaging by improving scribble-based segmentation, offering a practical solution for healthcare applications, though it is incremental in the context of weakly supervised methods.
The paper tackles the problem of medical image segmentation with limited scribble annotations by proposing HELPNet, which integrates hierarchical perturbations, entropy-guided pseudo-labels, and structural priors, achieving performance comparable to fully supervised methods on datasets like ACDC, MSCMRseg, and CHAOS.
Creating fully annotated labels for medical image segmentation is prohibitively time-intensive and costly, emphasizing the necessity for innovative approaches that minimize reliance on detailed annotations. Scribble annotations offer a more cost-effective alternative, significantly reducing the expenses associated with full annotations. However, scribble annotations offer limited and imprecise information, failing to capture the detailed structural and boundary characteristics necessary for accurate organ delineation. To address these challenges, we propose HELPNet, a novel scribble-based weakly supervised segmentation framework, designed to bridge the gap between annotation efficiency and segmentation performance. HELPNet integrates three modules. The Hierarchical perturbations consistency (HPC) module enhances feature learning by employing density-controlled jigsaw perturbations across global, local, and focal views, enabling robust modeling of multi-scale structural representations. Building on this, the Entropy-guided pseudo-label (EGPL) module evaluates the confidence of segmentation predictions using entropy, generating high-quality pseudo-labels. Finally, the structural prior refinement (SPR) module incorporates connectivity and bounded priors to enhance the precision and reliability and pseudo-labels. Experimental results on three public datasets ACDC, MSCMRseg, and CHAOS show that HELPNet significantly outperforms state-of-the-art methods for scribble-based weakly supervised segmentation and achieves performance comparable to fully supervised methods. The code is available at https://github.com/IPMI-NWU/HELPNet.