Projects

Annotation software

We are involved in two annotation frameworks for microscopy images.

EXACT

A web-based collaboration toolset for algorithm-aided annotation of images with annotation version control. Paper · GitHub

SlideRunner

A PyQt-based open source tool for massive cell annotations. Paper · GitHub

Datasets

We strongly believe in the importance of democratizing data access in order to drive advancements in our research domain. We have made the following data sets available:

CATCH

A tumor tissue segmentation dataset comprising 350 whole slide images of seven different canine cutaneous tumors, complemented by 12,424 polygon annotations for 13 histologic classes. Paper · GitHub · Data

MITOS_WSI_CMC

The currently largest dataset for breast cancer annotated on complete whole slide images. It contains 21 cases of canine breast cancer, labeled as the consensus of three experts. Paper · GitHub · Data

MITOS_WSI_CCMCT

The largest dataset for mitosis on canine cutaneous mast cell tumor, and the largest dataset with verified mitotic figures (around 44k cells). Each cell has been labeled as a consensus of three pathologists. Paper · GitHub · Data

TUPAC16_AL

A set of alternative labels for the TUPAC16 challenge auxiliary mitosis dataset. Our research paper shows that it's more complete and has a lower level of label noise. Paper · GitHub

EIPH

An inter-species cell detection dataset aimed at detecting pulmonary hemosiderophages in equine, human, and feline specimens. Paper · GitHub · Data

Pan-tumor T-lymphocyte detection dataset

An immunohistochemistry dataset stained positive for CD3. It comprises 92 ROIs from four tumor indications with cell-level annotations for CD3+ cells, tumor cells, and other cells. Paper · Data

Multi-scanner SCC dataset

A subset of the CATCH dataset digitized with five different scanning systems. The images provide local correspondences useful for domain shift experiments and WSI registration. Paper · GitHub · Data

MIDOG++

Extends the MIDOG 2021 and MIDOG 2022 challenge datasets. It comprises 503 ROIs of seven different tumor types with variable morphologic, laboratory, and scanner origins, with a total of 11,937 annotated mitotic figures. Paper · GitHub · Data

QuiltCleaner

A representative selection of impurity annotations for images from the QUILT-1M dataset, complemented with predictions for the remainder of the dataset. Paper · GitHub

Microscopy machine learning approaches

Our mission is not to improve insignificantly over previous work, but to define new and interesting approaches that lead to new insights in the field of microscopy and pathology. Some of our most recent work:

• Jonas Ammeling: Attention-based Multiple Instance Learning for Survival Prediction on Lung Cancer Tissue Microarrays
• Jonathan Ganz: Deep Learning-based Automatic Assessment of AgNOR-scores in Histopathology Images
• Marc Aubreville: Deep Learning-based Subtyping of Atypical and Normal Mitoses using a Hierarchical Anchor-free Object Detector
• Frauke Wilm: Mind the Gap: Scanner-induced domain shifts pose challenges for representation learning in histopathology

Challenges

We organized the Mitosis Domain Generalization Challenge (MIDOG) at MICCAI in 2021 and 2022. For both challenges, we published a challenge report paper in the journal Medical Image Analysis:

• Aubreville, M., Stathonikos, N., Bertram, C. A., Klopfleisch, R., Ter Hoeve, N., Ciompi, F., … & Breininger, K. (2023). Mitosis domain generalization in histopathology images — The MIDOG challenge. Medical Image Analysis, 84, 102699.
• Aubreville, Marc, et al. "Domain generalization across tumor types, laboratories, and species — Insights from the 2022 edition of the Mitosis Domain Generalization Challenge." Medical Image Analysis 94 (2024): 103155. Paper