The post-challenge journal paper is now published in IEEE Transcations on Medical Imaging and can be accessed from the following link-
Nuclear segmentation in digital microscopic tissue images could enable the extraction of high-quality features for nuclear morphometrics and other analysis in computational pathology. Techniques that accurately segment nuclei in diverse tissue images spanning a range of patients, organs, and disease states, can significantly contribute to the development of clinical and medical research software. Once accurately segmented, nuclear morphometric and appearance features such as nuclei density, nucleus-to-cytoplasm ratio, average size, and pleomorphism can be used to assess not only cancer grades but also for predicting treatment effectiveness. Identifying different types of nuclei based on their segmentation can also yield information about gland shapes, which, for example, is important for cancer grading.
This challenge will showcase the best nuclei segmentation techniques that will work on a diverse set of H&E stained histology images obtained from different hospitals spanning multiple patients and organs. This will enable the training and testing of readily usable (or generalized) nuclear segmentation softwares.
MoNuSeg is an official satellite event of MICCAI 2018.
To participate in this challenge please click join. If you do not have an account with grand-challenge.org then sign up for one, otherwise just sign with your registered credentials. If you are participating in a team of two or more then please use only one ID to register for the contest.
Participants are required to submit a 2 page manuscript (by July 15th) detailing their proposed nuclei segmentation algorithm. During the challenge session at MICCAI 2018, participants will present their methods to the live audience. Final results will be presented by the organizers. All algorithms will be included in a journal publication and each participating team will be allowed two co-authorships at most.
The two page manuscript can be sent at firstname.lastname@example.org