SiNuS: A Comprehensive Dataset for Singular Nuclei Segmentation for HER2 Grading of Breast Cancer

This dataset provides clinically validated annotations for singular nuclei segmentation in Dual-ISH breast cancer images (20×), essential for automated HER2 grading according to ASCO/CAP guidelines. Three expert pathologists independently annotated 39 image patches, resulting in 1,856 inclusive and 1,284 exclusive singular nuclei from 21 patients. The dataset includes binary, multicolor, and boundary masks, offering the first public resource dedicated to developing and evaluating singular nuclei segmentation and HER2 grading methods in digital pathology. Please cite the following paper if you use this dataset: Hossain, Md Shakhawat, Md Sahilur Rahman, and Munim Ahmed. "SiNuS: A Comprehensive Dataset for Singular Nuclei Segmentation for HER2 Grading of Breast Cancer." Data in Brief (2026): 112934.