Assessing the optimal technique to determine reference SUV for tissue to reference ratios using F18-DCFPyL in Whole Body PET-MR imaging
Bibliographic record
Abstract
2005 Objectives: Prostate specific membrane antigen (PSMA) is highly expressed in most prostate cancer cells [1]. New PET radiolabeled tracers like F18-DCFPyL, bind to PSMA and show promise in the diagnosis, staging and longitudinal assessment of response to therapeutic interventions [1]. Degree of tracer uptake at tumor sites is semi-quantitatively assessed in comparison to reference tissues. This decreases variability in interpretation and is the basis for the newly proposed molecular imaging classification system for prostate cancer (‘PROMISE’) [2]. Currently, there is a lack of data on the variability in the reference tissues used to assess degree of uptake, namely blood pool activity, liver and parotid glands. The purpose of the current study was to determine the variability in SUV measurement for reference tissues on F18-DCFPyL PET-MR imaging. Methods: Whole body PET-MR images from 5 male patients injected with F18-DCFPyL (325MBq ±10%) were analyzed retrospectively. At 120mins (±10mins) post injection, images were acquired from vertex to mid-thigh using 5 bed positions at 4mins each. PET-MR data were reconstructed using the 2-point Dixon MR attenuation correction, HDPET (Point-spread function), 3 iterations with 21 subsets, 172x172 matrix, and absolute scatter correction. In addition to no post-reconstruction smoothing (All-Pass), Gaussian filters at different full width half maximum (4, 5, 6 cm) were applied (4 sets of images), in order to evaluate the effect of image noise on the analyses. Sphere VOIs were drawn over the right parotid gland (1cm3 ±0.5cm3), the left ventricle cavity as a representative of blood pool (3cm3 ±0.5cm3), and the caudal aspect of the right liver lobe (3cm3 ±0.5cm3) on all 4 sets of images. SUVmax, SUVmin, SUVmean as well as standard deviations were calculated for all VOIs. Three out of the 5 patients underwent post-treatment imaging with F18-DCFPyL within a 5-7 month timeframe. The same reconstruction methods and VOIs were applied to the post-treatment image sets. Coefficient of variation (CV) was calculated for all SUV values in each reference region and in all reconstructed image sets. To determine test-retest variability in the 3 patients with pre and post treatment imaging, mean % differences (% ± SD) were calculated in each reference tissue VOI. Results: Results showed that the SUVmean for the VOI in the right lower lobe of the liver (CV ranging from 0.11% to 0.51%) and left ventricle (CV ranging from 0.23% to 1.00%) were relatively more stable for all reconstruction methods on the PET-MR images than the VOI in the right parotid gland (CV ranging from 0.19% to 8.44%). As for test-retest variability, the absolute SUVmean difference was less than 10% for the VOIs in the right lower lobe of the liver (2.55% ± 2.4) and left ventricle (4.79% ± 1.96) and greater than 10% for the right parotid gland (15.35 % ± 12.53). Conclusions: Test-retest for tissue VOIs in the right lower lobe of the liver and the left ventricle showed a more reliable finding than the right parotid gland. Both liver and blood pool VOIs produced less variability than the parotid gland. These findings warrant consideration in future studies applying semi-quantitative methods for longitudinal assessment of lesions over time using F18- DCFPyL in PET-MR imaging.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.006 | 0.011 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".