We are so glad to announce that the MIDOG 2021 challenge was accepted to be officially held together with MICCAI 2021 in Straßbourg/France.
We thank all reviewers for their valuable comments, which helped to improve the challenge design significantly.
I work as a software developer for Texas A&M Veterinary Medical Diagnostic Laboratory. Here at TVMDL we are working with AWS to develop ML models for both mitotic field as well as chronic waste disease detection(s). We are beginning the data labeling phase, can you suggest any open source software tools? Feel free to reach out to me and or TVMDL.
Thanks for you time.
Dear Steve,
thanks for getting in touch. In fact, we have been (and still are) developing open source software tools for annotation.
– If you are looking into having a standalone solution with minimal installation effort, I might recommend SlideRunner (written and maintained by me). https://github.com/DeepPathology/SlideRunner
– If you are looking for a multi-user solution, we also got you covered. My colleague Christian Marzahl recently released EXACT, a web-based labeling software that even includes Slide co-registration and other nice features. https://github.com/ChristianMarzahl/Exact.
Of course, there are also alternatives from other developer groups. You might want to look into cytomine (https://cytomine.be), which is also open source, and also web-based. We found it to be a bit high effort in terms of
mitosis (which should be from our perspective a single-click interaction to do).
And lastly, another standalone solution is QuPath (https://qupath.github.io), also with a large developer community.
It goes without saying that we’ve been using SlideRunner for our mitotic figure data sets, as it’s a solution meant for that purpose – and our pathologists are really fast with it. 🙂
Hope it helps. Best regards from Germany,
Marc