indi
indi > RE: ADHD-200 Global Competition Scoring Policy
Apr 6, 2011 02:04 PM | Michael Milham
RE: ADHD-200 Global Competition Scoring Policy
Paolo,
All test datasets will have both structural and resting fMRI images (1 of each). In the NYU dataset included in the training release, that were 4-5 subjects who had aborted rest scans, but reasonable structurals...we included them in the release to provide folks with maximal data.
Participants on the competition can choose to base their classifiers on rest scans, structural, or some combo of the two.
Hope that helps,
M
Originally posted by Paolo Avesani:
All test datasets will have both structural and resting fMRI images (1 of each). In the NYU dataset included in the training release, that were 4-5 subjects who had aborted rest scans, but reasonable structurals...we included them in the release to provide folks with maximal data.
Participants on the competition can choose to base their classifiers on rest scans, structural, or some combo of the two.
Hope that helps,
M
Originally posted by Paolo Avesani:
A question related to release of test data. I noticed, for example
in the NYU dataset, that not all subjects have at least one
recording of resting state. If this condition holds for test data
it means that I need to be in the position to classify a subject
without relying on functional data. It would be really important to
clarify this point. Will the test data include subjects with at
least one session of resting state recording?
Threaded View
| Title | Author | Date |
|---|---|---|
| Michael Milham | Apr 5, 2011 | |
| Paolo Avesani | Apr 6, 2011 | |
| Michael Milham | Apr 6, 2011 | |