Expert-Annotated Single-Cell Thyroid Cytology Dataset (SC-TC)

This dataset contains an expert-annotated collection of single-cell images for thyroid cancer diagnosis. It was curated from whole-slide images provided by medical specialists from Vietnam 108 Military Hospital and carefully annotated at the single-cell level to capture diverse morphological patterns of thyroid cells. Each cell image is associated with multiple binary labels describing nuclear characteristics that are relevant for cancer diagnosis. In total, the dataset includes 3,419 single-cell images with 9 expert-defined labels, making it one of the first publicly available resources for studying imbalanced multi-label classification in thyroid cytology. This dataset is intended to support research in deep learning, medical image analysis, and computer-aided cancer diagnosis, and it is accompanied by deep learning benchmark results presented in our paper “A Novel Expert-Annotated Single-Cell Dataset for Thyroid Cancer Diagnosis with Deep Learning Benchmarks”.