Data-driven respiratory motion correction of cardiac SPECT using a convolutional neural network
Bibliographic record
Abstract
Respiratory motion leading to degradation of image quality can reduce the diagnostic utility of SPECT myocardial perfusion imaging (MPI) studies. While retrospective motion compensation (MC) is possible with listmode acquisitions, many methods for doing so require additional equipment or time-consuming analysis, making it difficult to incorporate in a routine clinical workflow. This work presents an AI-based tool for motion detection in cardiac pinhole SPECT scans that does not need image reconstruction. This tool makes use of a convolutional neural network trained on respiratory-gated MPI studies from 90 patients (179 scans) supplemented with simulated scans using the NCAT digital torso phantom. Data are binned into respiratory gates based on the amplitude of detected count rate in 100 ms frames, and gated projection data are processed to determine if motion compensation is necessary based on a motion amplitude threshold of 12 mm. A beta version of this tool is currently in use at the University of Ottawa Heart Institute where it was installed on the Windows workstation used for processing scans. The AI tool provides feedback within 10 seconds, allowing technologists to immediately flag studies for further follow-up with minimal disruption to workflow. On an initial evaluation of 21 studies the tool shows a 73.3% success rate in identifying scans that do not require additional MC processing. In the first week of usage, seven studies were flagged including four showing a visual improvement with MC. Using feedback from clinical technologists and initial findings, modifications to the AI tool are still being implemented.
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 | 0.001 | 0.001 |
| 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 | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| 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".