Detection and severity classification of extracardiac interference in <sup>82</sup>Rb PET myocardial perfusion imaging
Bibliographic record
Abstract
PURPOSE: Myocardial perfusion imaging (MPI) is used for diagnosis and prognosis of coronary artery disease. When MPI studies are performed with positron emission tomography (PET) and the radioactive tracer rubidium-82 chloride ((82)Rb), a small but non-negligible fraction of studies (∼10%) suffer from extracardiac interference: high levels of tracer uptake in structures adjacent to the heart which mask the true cardiac tracer uptake. At present, there are no clinically available options for automated detection or correction of this problem. This work presents an algorithm that detects and classifies the severity of extracardiac interference in (82)Rb PET MPI images and reports the accuracy and failure rate of the method. METHODS: A set of 200 (82)Rb PET MPI images were reviewed by a trained nuclear cardiologist and interference severity reported on a four-class scale, from absent to severe. An automated algorithm was developed that compares uptake at the external border of the myocardium to three thresholds, separating the four interference severity classes. A minimum area of interference was required, and the search region was limited to that facing the stomach wall and spleen. Maximizing concordance (Cohen's Kappa) and minimizing failure rate for the set of 200 clinician-read images were used to find the optimal population-based constants defining search limit and minimum area parameters and the thresholds for the algorithm. Tenfold stratified cross-validation was used to find optimal thresholds and report accuracy measures (sensitivity, specificity, and Kappa). RESULTS: The algorithm was capable of detecting interference with a mean [95% confidence interval] sensitivity/specificity/Kappa of 0.97 [0.94, 1.00]/0.82 [0.66, 0.98]/0.79 [0.65, 0.92], and a failure rate of 1.0% ± 0.2%. The four-class overall Kappa was 0.72 [0.64, 0.81]. Separation of mild versus moderate-or-greater interference was performed with good accuracy (sensitivity/specificity/Kappa = 0.92 [0.86, 0.99]/0.86 [0.71, 1.00]/0.78 [0.64, 0.92]), while separation of moderate versus severe interference severity classes showed reduced sensitivity/Kappa but little change in specificity (sensitivity/specificity/Kappa = 0.83 [0.77, 0.88]/0.82 [0.77, 0.88]/0.65 [0.60, 0.70]). Specificity was greater than sensitivity for all interference classes. Algorithm execution time was <1 min. CONCLUSIONS: The algorithm produced here has a low failure rate and high accuracy for detection of extracardiac interference in (82)Rb PET MPI scans. It provides a fast, reliable, automated method for assessing severity of extracardiac interference.
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 | 0.007 | 0.016 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 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".