Code and MIDOG2022 Dataset for "Cross-Scale Hypergraph Neural Networks with Inter-Intra Constraints for Mitosis Detection"

This Zenodo entry provides the source code and the MIDOG2022 dataset used in our study titled "Cross-Scale Hypergraph Neural Networks with Inter-Intra Constraints for Mitosis Detection". Our method introduces a novel hypergraph-based model incorporating both inter- and intra-scale constraints for accurate mitosis detection on whole slide images (WSIs). To ensure reproducibility, we release: The complete source code for training and evaluation the MIDOG2022 dataset(Getting the data)and their Zenodo DOI is 10.5281/zenodo.6547151 Please note that we have uploaded the MIDOG2022 dataset using split archive compression. To extract the full dataset, you need to download all of the following files: MIDOG2022 Dataset.z01 to MIDOG2022 Dataset.z06 MIDOG2022 Dataset.zip Make sure all files are placed in the same directory, then open the MIDOG2022 Dataset.zip file using a standard archive tool to extract the complete dataset. Additionally, please be aware that we have uploaded the PNG version of the MIDOG2022 dataset due to Zenodo’s file size limitations.If you wish to obtain the original .tiff files, you can download them from the official website: Getting the data This release is compliant with the reproducibility requirements and allows researchers to replicate and build upon our work.