MétaCan
Menu
← Back to cohort
Record W4256410048 · doi:10.22215/etd/2015-11070

Extra-Cardiac Interference in Myocardial Perfusion Imaging with Rubidium-82 and Positron Emission Tomography

2015· dissertation· en· W4256410048 on OpenAlexafffund
Elizabeth Orton

Bibliographic record

Venuenot available
Typedissertation
Languageen
FieldMedicine
TopicMedical Imaging Techniques and Applications
Canadian institutionsCarleton UniversityOttawa Hospital
FundersNatural Sciences and Engineering Research Council of Canada
KeywordsPositron emission tomographyCardiac PETMyocardial perfusion imagingNuclear medicinePerfusionConcordanceArtificial intelligenceComputer scienceMedicineAlgorithmCardiologyInternal medicine

Abstract

fetched live from OpenAlex

Interpretation of myocardial perfusion images, produced with blood flow tracer rubidium-82 chloride ( 82 Rb) and positron emission tomography (PET), can be affected when high tracer uptake in extra-cardiac organs adjacent to the heart (stomach) interferes with the myocardium.Since extra-cardiac organs are physically spatially distinct from the myocardium, extra-cardiac interference (ECI) in 82 Rb PET images arises from limited spatial resolution, and cardiac and respiratory motion.This thesis aims to provide automated methods that detect and correct ECI.Three algorithms were developed to fulfill these aims: the first detects and ranks severity of ECI, the second attempts ECI correction based on factor analysis of dynamic image series, and the third corrects ECI with a 1D convolution-based method.All algorithms were developed, implemented and evaluated based on sets of clinical images.The detection and severity classification (DSC) algorithm was developed based on concordance of a 200 image dataset with clinical interpretation.It detected ECI with high accuracy (97% sensitivity and 82% specificity), low failure rate (<1%) and short execution time (<7s).The algorithm was used to estimate prevalence of ECI in a 4920image dataset and to determine if simple modifications to image processing protocols could reduce ECI prevalence and/or severity.While reduced filtering showed the most promise, none of the available modifications eliminated ECI in the majority of images.Factor analysis of dynamic image series uses differences in the temporal behaviour of tracer uptake in the myocardium compared to that in the extra-cardiac organ to separate the two structures.Variations of this approach, applied to 82 Rb PET images, i were not able to simultaneously correct images with ECI of all severities and avoid reducing myocardial intensity in images without interference prior to correction.The 1D convolution-based correction algorithm modeled the image point spread function, including the effects of motion, as a 1D Gaussian and the underlying myocardial and stomach tracer uptake as simple 1D rectangular functions.This algorithm corrected images with ECI of all severities without reducing myocardial intensity in ECI-free images and >90% of scans examined showed visually acceptable correction.The convolution-based correction algorithm shows promise as a softwarebased ECI correction.The duration of my degree necessitates I thank many people, all of whom have been a part of my PhD-years.I have the most reason to thank my supervisor, Glenn Wells.His patience, guidance and support have not only helped me arrive at this point but have both directly and indirectly provided me a new set of skills, with which I can approach future intellectual pursuits.A number of other Carleton medical physics faculty have provided me with particular support that I would like to acknowledge.Thank you to Brenda Clark and Joanna Cygler for

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 imitation

Not 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.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.001
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Bench or experimental · Consensus signal: Bench or experimental
GenreCandidate signal: Empirical · Consensus signal: none
Teacher disagreement score0.004
Threshold uncertainty score0.015

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0010.001
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.001
Science and technology studies0.0000.000
Scholarly communication0.0010.000
Open science0.0000.001
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0040.002

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.

Opus teacher head0.012
GPT teacher head0.301
Teacher spread0.290 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designBench or experimental
Domainnot available
GenreEmpirical

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".

Quick stats

Citations0
Published2015
Admission routes2
Has abstractyes

Explore more

Same topicMedical Imaging Techniques and Applications→French-language works237,207→