LGDUnet: a dual-branch network for weakly supervised gland segmentation in H&E stained histopathological images.

Glandular segmentation plays a critical role in the pathological analysis of hematoxylin and eosin (H&E)-stained images, particularly for the diagnosis and treatment of breast and colorectal cancers. Accurate segmentation of glandular structures provides essential support for lesion detection and pathological assessment. Traditional fully supervised learning methods typically require a substantial amount of labeled data, which is often challenging to obtain in the medical field. To address this issue, we propose a novel weakly supervised segmentation method that integrates pseudo-labeling with a noise-consistency loss, achieving performance comparable to several state-of-the-art architectures trained under full supervision. The proposed network, termed LGDUnet, adopts a dual-branch architecture consisting of global and local branches. The global branch focuses on the overall structure of the glands, while the local branch emphasizes the capture of fine edge features. The encoder employs a combination of ResHorBlock and WSAB to enhance feature extraction through higher-order interactions and weighted spatial attention. The decoder employs the Dual Self-Attention Transpose (DST) structure, which enhances reconstruction accuracy through a dual self-attention mechanism. Skip connections are implemented using the Dual Attention Transformer (DAT) for encoder feature transformation, further improving the efficacy of feature propagation. We conducted comprehensive comparative and ablation experiments on the benchmark dataset GlaS from the MICCAI 2015 Challenge, our self-constructed breast cancer dataset Tubule of Breast Cancer (TBC), and the Colorectal Cancer Gland Dataset (CGD). Experimental results demonstrate that LGDUnet achieves superior performance in glandular segmentation tasks, validating its effectiveness under the weakly supervised learning framework.