fidoptLung250M-4B
Chest CT of one patient, inspiration and expiration or baseline and follow-up, with manual landmarks (Lung250M-4B validation pairs).
Read before comparing (2)
- Truth. The dataset's manual landmarks, 100 a pair. The masks are the dataset's lung masks.
- Overlap. Pairs 114 to 123 are the scans of L2R_NLST's ten validation pairs, here on another grid and with manual landmarks instead of automatic keypoints.
No method has a result on every pair of this dataset yet.
| Accuracy | Smoothness | Cost | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| # | Method | TRE mm | Change | rTRE | SDlogJ | Folds % | Time s | RAM GB | GPU GB | Images |
| No registration | 13.11 ±2.96 | 0.0282 ±0.0069 | 0.000 | 0.00 | ||||||
TRE: distance between matching landmarks. rTRE: the TRE over the length of the image's diagonal, to compare datasets of other sizes and units. SDlogJ: how irregular the deformation is, 0 is perfectly smooth. Folds: share of the image where the deformation folds over itself.
Pair by pair
Click a score to look at that result: the fixed image, and over it the moving image after registration, with the annotations. Underlined scores have their image; for the others you see the pair without the result. A green cell gained against no registration on that pair, a red one lost.
About the data
- Test pairs
- The 12 validation pairs with manual landmarks that no other dataset here holds, two from TCIA-Ventilation and ten from TCIA-NLST. Of the 27 pairs with manual landmarks, 054 to 056 are L2R_LungCT's validation pairs, and 002, 008 (EMPIRE10) and 104 to 113 (DIR-Lab COPDgene) come without their images. The 3 development cases are training pairs, whose correspondences are automatic.
- Truth
- landmarks by experts the dataset's manual landmarks, 100 a pair
- Licence
- CC BY 4.0, except its EMPIRE10 and DIR-Lab COPDgene cases
3 pairs for tuning, never tested
case_012, case_013, case_014