SPIE-AAPM-NCI Prostate MR Classification Challenge
SPIE, along with the support of the American Association of Physicists in Medicine (AAPM) and the National Cancer Institute (NCI), will conduct a “Grand Challenge” on quantitative image analysis methods for the diagnostic classification of clinically significant prostate lesions. As part of the 2017 SPIE Medical Imaging Symposium, the PROSTATEx Challenge will provide a unique opportunity for participants to compare their algorithms with those of others from academia, industry, and government in a structured, direct way using the same data sets.
Please note the updated timeline below!
Release date of training set cases with truth: 21 Nov 2016
Release date of test set cases without truth: 12 Dec 2016
Submission date for participants’ test set classification output: 16 Jan 2017
Challenge results released to participants: 20 Jan 2017
SPIE Medical Imaging Symposium: 13-16 Feb 2017
For more information please refer to the PROSTATEx Challenge Format below.
A special session at the 2017 SPIE Medical Imaging Symposium will focus on the PROSTATEx Challenge, and the two top-performing participants will present their methods during this session and receive free conference registration and publicity. Challenge participants who submit test set classification results by the 16 January 2017 deadline will be invited to present a poster and demonstrate their algorithm at the Live Demonstration CAD Workshop during the Symposium.
Participants are encouraged to submit their work to the SPIE CAD conference proceedings volume as well as for peer review to the SPIE’s Journal of Medical Imaging.
PROSTATEx Challenge Format
The database for this challenge will contain approximately 350 MRI cases, each from a distinct patient at a single exam time. About 60% of the cases will serve as the training set and the remaining 40% as the test set with each case consisting of four sets of DICOM-formatted MRI scan data: T2-weighted images (transaxial and sagittal, DICOM format), Ktrans images (computed from dynamic contrast-enhanced (DCE) images, mdh format), and apparent diffusion coefficient (ADC) images (computed from diffusion-weighted (DWI) imaging, DICOM format). Each case will contain at least one prostate lesion with biopsy-proven malignancy status or with imaging findings with sufficiently low suspicion of clinical significance. Lesion locations will be provided, but Gleason scores and PI_RADS scores will not be released with this challenge, since they will become the focus of a follow-up challenge to be held in conjunction with the 2017 AAPM Annual Meeting in July 2017.
Participants may use the training set cases in any manner they would like for the purpose of training their systems; there will be no restrictions on the use of the data or the advice sought from local experts for training purposes. The test set cases, however, are to be manipulated, processed, and analyzed without human intervention (although human-supervised delineation of the prostate gland border (not the lesion margin) will be allowed, if the operation of a computerized system requires such input).
Participants are free to download the training set and, subsequently, the test set when these datasets become available. It is important to note that once participants submit their test set classification output to the challenge organizers, they will be considered fully vested in the Challenge, and their performance results (without links to the identity of the participant) will become part of any presentations or publications derived from the Challenge at the discretion of the organizers. The truth associataed with the test set cases is expected to be made publicly available after the publication of the PROSTATEx Challenge and its planned follow-up challenge.
Participation in the PROSTATEx Challenge acknowledges the educational, friendly competition, and community-building nature of this challenge and commits to conduct consistent with this spirit for the advancement of the medical imaging research community.
See the Guest Editorial: LUNGx Challenge for computerized lung nodule classification: reflections and lessons learned for a discussion of lessons learned from the 2015 LUNGx Challenge, also sponsored by SPIE, AAPM,a nd NCI.
The data used in this challenge was prepared by
The Cancer Imaging Archive (TCIA).
You can begin the challenge here: SPIE Challenges
Organizers and Major Contributors:
- Sam Armato, The Univ. of Chicago (United States)
- Henkjan Huisman, Radboud Univ. (Nehterlands)
- Karen Drukker, The Univ. of Chicago (United States)
- Keyvan Farahani, National Institutes of Health (United States)
- Nicholas Petrick, U.S. Food and Drug Administration (United States)
- Maryellen Giger, The Univ. of Chicago (United States)
- Lubomir Hadjiiski, Univ. of Michigan (United States)
- Jayashree Kalpathy-Cramer, Harvard Univ. (United States)
- Artem Mamonov, Harvard Univ. (United States)
- Robert Nordstrom, National Institutes of Health (United States)
- Justin Kirby, National Institutes of Health (United States)
- George Redmond, National Institutes of Health (United States)
- Shayna Knazik, AAPM (United States)
- Diane Cline, SPIE
- Robbine Waters, SPIE (RobbineW@spie.org)