Quantifying Uncertainty in SPECT Imaging Using a Gamma Statistical Model with Spatial Covariance
Bibliographic record
Abstract
PurposeUncertainty assessment improves radiopharmaceutical dosing accuracy, aiding clinicians in predicting therapeutic outcomes and minimizing side effects. It ensures that scanner performance aligns with design specifications through routine quality checks on sensitivity, bias, and 3-D capabilities. Building on research that normalized PET data with the Gamma model, we aim to extend this method to SPECT, analyzing statistical distributions and spatial covariance in reconstructed images to refine noise recalibration and enhance uncertainty estimates.MethodsThe method involves two main steps: data analysis within the ROI and simulated data generation. Initially, voxel-level shape and scale parameters in the ROI are estimated through multiple scans or by re-binning List-mode data from a single scan to adjust for changes due to radioactive decay and biological clearance. We used 5 images to normalize data and construct a spatial autoregressive matrix, which scales simulated Gaussian white noise accordingly. The white-noise variance equals the mean voxel value, consistent with Poisson distribution. These images help calculate the ROI’s relative error, defined as the standard deviation divided by the mean value. Validation involves using the XCAT phantom and SIMIND software to simulate a patient scenario with a liver tumor post 6.85 GBq 177Lu-DOTATATE injection. OSEM reconstructions are performed with 10 to 30 iterations across 6 subsets.Results and DiscussionThe covariance from 600 noise realization images guided the selection of a 5-order neighbor model for the spatial autoregressive matrix, optimizing accuracy and simplicity. Results showed symmetric covariance, allowing us to limit optimization to 5 coefficients and speed up computations. This model application significantly reduced autocorrelation and showed that relative error in total counts across the ROI or tumor area increases with the number of OSEM reconstruction iterations.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 teacher head, 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".