fidoptACRIN 6668, CT to FDG-PET
CT and FDG-PET from one PET/CT session of a patient with non-small-cell lung cancer in the ACRIN 6668 trial: the CT is the source, the PET in body-weight SUV on its native grid the target.
11 test caseswhole bodysource CT, target PET
Read before comparing (6)
- Pairs. A case is a CT and an attenuation-corrected FDG-PET from one PET/CT session, one session a patient. prepare.py checks that both series share their FrameOfReferenceUID and pins their SeriesInstanceUIDs. Nothing is registered: the CT is resampled onto the PET grid with a cubic B-spline, air outside the CT, and rounded to whole Hounsfield units.
- Target. Body-weight SUV, computed from the PET's rescale values and its DICOM dose, weight and decay information by AutoPET_CTPET_Syn.suv_factor. The mask is the body in the shared field of view, made by AutoPET_CTPET_Syn.body.
- Limits. The PET and CT have different acquisition durations, and scanner alignment does not remove the respiratory motion between them. Tracer uptake is not determined by CT anatomy alone.
- Range. A dynamic range of 20 SUV gives PSNR and SSIM the same scale as the corpus's other FDG-PET synthesis datasets; it clips neither the target nor its errors.
- Baseline. On the 11 test cases a constant at the target's median inside the mask scores, as the median over the cases and its range, an MAE of 0.53 (0.33 to 0.65) SUV, a PSNR of 23.8 (13.7 to 28.3) dB and an SSIM of 0.74 (0.70 to 0.81).
- Overlap. The ACRIN-NSCLC-FDG-PET patients of every descriptor were checked. SAROSBodyCT holds 23: the 11 in this pool are held out here, the 12 others are not eligible. TotalSegMR's test image s0031 may be an MR series of this collection from a patient its notes do not name: an MR, neither this dataset's CT nor its PET.
No method has a result on every case of this dataset yet.
Case by case
Click a score to look at that result: the target image, and over it the method's synthetic image. Underlined scores have their image; for the others you see the case without the result.
About the data
The first test caseClick to look at every case.
- Test cases
- 11 test patients drawn with seed 2026, stratified by scanner model, and ten development patients, all distinct. The pool is the 129 patients with an eligible paired session; the 11 of them among SAROSBodyCT's test and development cases, whose CT SAROSBodyCT shows, are held out, leaving 118.
- Truth
- target image as released the attenuation-corrected FDG-PET of the same PET/CT session in body-weight SUV, aligned by the scanner frame
- Licence
- CC BY 3.0
10 cases for tuning, never tested
ACRIN-NSCLC-FDG-PET-039, ACRIN-NSCLC-FDG-PET-087, ACRIN-NSCLC-FDG-PET-129, ACRIN-NSCLC-FDG-PET-091, ACRIN-NSCLC-FDG-PET-232, ACRIN-NSCLC-FDG-PET-033, ACRIN-NSCLC-FDG-PET-096, ACRIN-NSCLC-FDG-PET-192, ACRIN-NSCLC-FDG-PET-143, ACRIN-NSCLC-FDG-PET-128