Answers are grounded in this paper and cite their sources.
Computational study demonstrates enhanced glomerulus segmentation in kidney pathology images, indicating potential to improve chronic kidney disease diagnosis.
Chronic kidney disease (CKD) is a health concern that necessitates early diagnosis and intervention. Accurate segmentation of glomeruli in kidney images is essential for effective assessment and diagnosis of CKD. Current segmentation algorithms face challenges in distinguishing glomeruli from surrounding tissues, which can hinder clinical decision-making. Experimental results demonstrate that KPIs-RSFNet consistently improves Dice coefficient in glomerulus segmentation tasks compared with baseline models. The proposed KPIs-RSFNet demonstrates effective glomerulus segmentation performance and may serve as a useful computational tool for CKD-related pathological image analysis. We first perform image pre- processing to enhance data diversity. KPIs-RSFNet then leverages the strengths of ResNet152 and Swin Transformer, utilizing an attention mechanism to effectively fuse their feature representations, which allows for better capture of the complex characteristics of glomeruli. Finally, an edge detection algorithm is applied to overlay the segmentation results onto the original images, creating an intuitive visualization effect.
No takes yet. Share an insight, caveat, or question.
Li et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: