Investigating the Roles of Reconstruction and the Self-Calibration Factor for 90Y SPECT/CT Image-based Dosimetry
Bibliographic record
Abstract
1017 Objectives: The local deposition, dose-point kernel convolution, and Monte Carlo methods are currently the three image-based dosimetry techniques for Yttrium-90 (90Y) radioembolization. Image-based, these dosimetry techniques are dependent on the method of image reconstructions performed. In regards to dosimetry, the different effects between subset and iteration numbers have not been investigated. 90Y SPECT/CT dosimetry further utilizes a self-calibration factor to equate counts per voxel to activity per voxel based on the rationale that all the activity is confined within any specified ROI. Consequently, we investigate the role of these factors on image-based dosimetry through the local deposition method. Methods: In this presented work, a single patient’s 90Y SPECT/CT image was retrospectively used to investigate the impact of different reconstruction parameters and self-calibration factors. This patient was treated with of 90Y glass microspheres. In this work, we performed a total of seven OSEM reconstructions with HybridRecon (v1.0 Hermes) with a varying number of subsets and iterations to an equivalent of 150 MLEM iterations (e.g. 5sub30iter). All reconstructions were either performed with 8.5mm Gaussian post-filtering or reconstructed with OSEM MAP with median root prior (MRP) and a Bayesian weight of .3. A recommend clinical reconstruction was further performed with 5 iterations, 15 iterations, and a .4 cm-1 Butteworth filter (Clinical). These reconstructions were corrected for collimator scatter, scatter, had resolution recovery, and attenuation corrections based on CT. Contours of the liver, lung, and body were drawn using MIM Maestro v6.6. All processing of data and the local deposition method was developed in python code. Voxel densities were calculated using a scanner specific linear lookup table based on electron density phantom scans. The local deposition used the resulting voxel densities from the CT and interpolated SPECT voxels to perform absorbed dose calculations. For dose calculations, the 90Y half-life was set at 64.24 hours, used an average energy of .935 MeV per disintegration, and set the self-calibration factor dependent on the a specific contoured ROI (FOV, Body, and LiverLung). Due to an incomplete lung volume shown in the CT, an adjusted self-calibration method was additionally introduced that combined technetium macro-aggregated albumin based lung-shunting fraction with the same LiverLung contours. This method (TcMAA) allowed the liver and lung activities to be adjusted by a lung-shunt fraction resulting in two self-calibration factors specific to the liver and lung. Results: The range for liver and lung mean absorbed doses was 9.92 to 24.93 Gy and 2.84 to 8.08 Gy, respectively. These doses were dependent upon the combination of the self-calibration factor and type of reconstruction algorithm. Figure 1’s boxplots illustrate that the self-calibration factor had a bigger effect in determining the change in mean absorbed dose calculations, than reconstructions. Conclusions: The provided results illustrate that mean absorbed dose calculations to the target and healthy tissues can vary depending on the self-calibration factor and type of the reconstruction algorithms used. Consensus regarding which parameters to use is important for more accurate dosimetry.
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 | 0.003 | 0.009 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 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 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".