Cervix93 Image Reclassification Dataset: ASCUS, LSIL, HSIL, and Negative Labels for AI Research in Cytopathology

Cervix93 Image Reclassification Dataset: ASCUS, LSIL, HSIL, and Negative Labels for AI Cytopathology This dataset presents a pathologist-verified reclassification of cervical cytology images derived from the publicly available Cervix93 dataset. Each image corresponds to an Extended Depth of Field (EDF) microscopy frame from cervical smear slides. A qualified clinical pathologist manually assigned each image to one of four diagnostic categories following the Bethesda system: ASCUS – Atypical Squamous Cells of Undetermined Significance LSIL – Low-Grade Squamous Intraepithelial Lesion HSIL – High-Grade Squamous Intraepithelial Lesion Negative – No signs of intraepithelial lesion or malignancy The reclassified dataset is structured into four folders (one per diagnostic category), each containing the corresponding images. A companion CSV file is also provided, mapping each image filename to its assigned class label. 🔗 Original Dataset Attribution The images included in this dataset were originally published as part of the Cervix93 dataset by Parham et al. (2018). We provide this reclassified version to support reproducible research in medical image analysis and AI-assisted cervical cancer screening. 📄 Original Paper: https://arxiv.org/pdf/1811.09651 📂 Original Dataset Repository: https://github.com/parham-ap/cytology_dataset Please ensure that both this reclassification work and the original dataset authors are appropriately cited in any derivative research. 🔒 License This dataset is released under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license. ✅ Research and academic use is permitted 📌 Citation of the dataset authors is mandatory 🚫 Commercial use, redistribution, or modification is not allowed without prior written permission This dataset is intended solely for non-clinical, academic research. Diagnostic labels were assigned by a qualified pathologist but have not undergone regulatory or clinical validation.