MRI-guided PET Reconstruction with Adaptive Prior Strength: Study on Image Quality at Different Levels of Statistics
Bibliographic record
Abstract
Abstract BackgroundThe combination of magnetic resonance imaging (MRI)-anatomical information with positron emission tomography (PET) image reconstruction has been shown to improve PET image quality in terms of spatial resolution and image noise, especially in brain PET imaging. There are different approaches to combine MRI and PET available, being the use of a Bayesian framework the most extended. Generally, the strength of the prior is controlled by a hyperparameter that needs to be tuned depending on the acquired statistics/counts and the desired image quality in the resulting PET image. However, comparisons between methods is scant, and it is not clear how sensitive they are to the different levels of statistics that can be measured in a PET scan. MethodsIn this work we employed maximum a posteriori (MAP) reconstruction with MRI information to guide the PET reconstruction, and evaluated the performance of several prior models and optimization methods with a fixed and adaptive hyperparameter. Different simulated scenarios, and measured data using different radiotracers at different levels of statistics were employed for the evaluation. Comparisons in image quality and quanti cation between methods were performed in different brain cortical and sub-cortical regions. ResultsSimulated data showed that an adaptive hyperparameter consistently outperformed a fixed hyperparameter for every image reconstruction algorithm implemented. The best performance was achieved with a model combining the Bowsher prior weighted by similarity coefficients based on joint entropy between PET and MRI. One-step-late (OSL) and preconditioned gradient ascent (PGA) optimization methods performed similarly at any level of statistics and number of iterations, so long as the hyperparameter was adaptive. ConclusionsResults with simulated and measured data agreed that MAP reconstruction outperformed OSEM reconstruction, especially at low level of statistics, without any need of tuning. High resolution and low noise images were obtained using MAP reconstruction for 5{30 minutes scan times, showing negligible image quality difference for different radiotracers.
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 | 0.011 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| 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".