Read before comparing
- Nobody annotated these datasets by hand: their organ contours come from a segmentation program run on each image. A method that itself relies on that program is checked against its own view of the anatomy. These leaderboards are ranked apart for that reason.
- 1 dataset of this kind is shown below and not ranked on the official board: its licence does not allow a public leaderboard. The research board (Research, above the ranking) ranks it too.
- On UCSD-BMETS-Longitudinal, no method does better than no registration: as far as the annotations can tell, the pairs already match.
- On MSLesSeg, no method does better than no registration: as far as the annotations can tell, the pairs already match.
Provisional: no method has run on every dataset here yet. The methods that cover the most datasets come first, then the lower mean place on those.
1 place in that table, 1 is best2/10 cases done, not ranked yetn/a does not apply not run some cases failed
Under each dataset, the method's place in that dataset's table: 1 is best, the darker the better. The mean rank averages its places over the datasets.
The scores, dataset by dataset
Each score is followed by its change against no registration: green when the method did better, red when it did worse.
30 pairs · ordered on Dice and SDlogJ | # | Method | Dice | Change | SDlogJ |
|---|
| 1 | NiftyReg | 0.768 | −0.023 | 0.038 |
| 2 | Demons | 0.690 | −0.101 | 0.022 |
| No registration | 0.791 | | 0.000 |
10 pairs · ordered on Dice and SDlogJ 10 pairs · ordered on Dice and SDlogJ | # | Method | Dice | Change | SDlogJ |
|---|
| 1 | Demons | 0.727 | −0.073 | 0.034 |
| No registration | 0.800 | | 0.000 |
17 pairs · ordered on Dice and SDlogJ 10 pairs · ordered on TRE and SDlogJ 10 pairs · ordered on Dice and SDlogJ 10 pairs · ordered on Dice and SDlogJ 2 datasets without a ranked method yet
CT Colonography (ACRIN 6664), PanTrack.
fidopt