KDnuggets Home » News » 2016 » Jan » News, Features » Camelyon16 – Machine Learning Challenge in cancer detection ( 16:n02 )

Camelyon16 – Machine Learning Challenge in cancer detection


Camelyon16 challenge in conjugation with IEEE International Symposium on Biomedical Imaging is here! You have to design and develop a system which can detect and localize metastatic regions in whole slide microscopic images.



We are challenging strong groups in the Machine Learning and Image Analysis field to take on the difficult task of detecting metastases in microscopic images of lymph node tissue sections.

camelyon-16-grand-challenge

Camelyon16 challenge in conjugation with IEEE International Symposium on Biomedical Imaging (ISBI-2016)

Submission Deadline: April 1, 2016 | Challenge Event: April 13-16, 2016 | Location: Prague, Czech Republic

Overview:

Presence of metastasis in lymph nodes of cancer patients is a poor prognostic sign and prompts for more intense treatment. Automated detection of metastases is highly meaningful and holds a great promise to improve the diagnostic process.

We are providing two large datasets (400 microscopic high-resolution gigapixel images) from the Radboud University Medical Center (Nijmegen, the Netherlands), as well as the University Medical Center Utrecht (Utrecht, the Netherlands). This will be the first challenge using whole-slide images in histopathology giving the participants huge amounts of data to train their systems.

Technical Task:

Design and develop a system which can detect and localize metastatic regions in whole slide microscopic images. The following two strategies will be used to evaluate the performance of the algorithms:

  • Slide-based Evaluation: The merits of the algorithms will be assessed for discriminating between slides containing metastasis and normal slides. Receiver operating characteristic (ROC)¬†analysis at the slide level will be performed and the measure used for comparing the algorithms will be area under the ROC curve (AUC).
  • Lesion-based Evaluation: For the lesion-based evaluation, free-response receiver operating characteristic¬†(FROC) curve will be used. The FROC curve is defined as the plot of sensitivity versus the average number of false-positives per image.

As this challenge evaluates algorithms for both WSI classification and metastasis localization/detection, there will be two main leaderboards for comparing the algorithms.

Participation:

Interested participants are encouraged to get started immediately. To participate in the challenge and have your results visible on this website it is mandatory to submit a paper to the organizers explaining your algorithm

Please check out our website for further information about this challenge:

http://camelyon16.grand-challenge.org

Related:


Sign Up

By subscribing you accept KDnuggets Privacy Policy