Performance Improvements in HYPR-OSEM
Bibliographic record
Abstract
We describe methods which improve the performances in contrast recovery coefficient (CRC) versus noise trade-off and in convergence rate in CRC of the newly developed HYPR-OSEM algorithm. HYPR-OSEM is a reconstruction method which incorporates HighlY constrained back-PRojection (HYPR) de-noising directly within the widely used OSEM algorithm. 3 forms of HYPR-OSEM have been proposed. Previously, we have demonstrated that all forms of HYPR-OSEM can improve SNR without degrading accuracy in terms of resolution and contrast. However, slower convergence rate in CRC was observed from all forms of HYPR-OSEM. In this work, we investigated the effect of the filter kernel size used in the HYPR operator. Furthermore, we introduced the Iterative HYPR (IHYPR) operator as an effort to accelerate the convergence rate in CRC. Multiple independent noisy realizations of a simulated and an experimental contrast phantom with various sizes of hot and cold inserts were used for the evaluations. CRC vs voxel noise, image profile, and root-mean-squared error (RMSE) in CRC vs iteration were compared across standard and proposed reconstruction methods. Visual image quality assessment of a [11C]PK11195 patient scan was also conducted. It was observed that the noise reduction performance of HYPR-F(B)-OSEM is not very sensitive to the filter kernel size used in the HYPR operator, whereas better CRC vs noise trajectories and lower RMSE in CRC can be achieved by wider kernels for HYPR-AU-OSEM. On the other hand, the CRC convergence rate for HYPR-AU-OSEM becomes much slower with a wider kernel. When the IHYPR operator was introduced into the AU method (i.e. IHYPR-AU-OSEM), similar CRC convergence speed with respect to OSEM was attained without excessively degrading the CRC vs noise trajectories. In summary, the AU method has been determined to be the more effective form of HYPR-OSEM in terms of accuracy and precision, and IHYPR-AU-OSEM can achieve better CRC vs noise trajectories with similar convergence speed as compared to OSEM with or without a post reconstruction filter.
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 | 0.007 |
| 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 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 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".