Myocardial blood flow quantification by Rb-82 cardiac PET/CT: A detailed reproducibility study between two semi-automatic analysis programs.

Details

Ressource 1Download: BIB_03688DBECF5F.P001.pdf (2357.79 [Ko])
State: Public
Version: author
Serval ID
serval:BIB_03688DBECF5F
Type
Article: article from journal or magazin.
Collection
Publications
Institution
Title
Myocardial blood flow quantification by Rb-82 cardiac PET/CT: A detailed reproducibility study between two semi-automatic analysis programs.
Journal
Journal of nuclear cardiology
Author(s)
Dunet V., Klein R., Allenbach G., Renaud J., deKemp R.A., Prior J.O.
ISSN
1532-6551 (Electronic)
ISSN-L
1071-3581
Publication state
Published
Issued date
06/2016
Peer-reviewed
Oui
Volume
23
Number
3
Pages
499-510
Language
english
Notes
Publication types: Comparative Study ; Evaluation Studies ; Journal Article ; Validation Studies
Publication Status: ppublish
Abstract
Several analysis software packages for myocardial blood flow (MBF) quantification from cardiac PET studies exist, but they have not been compared using concordance analysis, which can characterize precision and bias separately. Reproducible measurements are needed for quantification to fully develop its clinical potential.
Fifty-one patients underwent dynamic Rb-82 PET at rest and during adenosine stress. Data were processed with PMOD and FlowQuant (Lortie model). MBF and myocardial flow reserve (MFR) polar maps were quantified and analyzed using a 17-segment model. Comparisons used Pearson's correlation ρ (measuring precision), Bland and Altman limit-of-agreement and Lin's concordance correlation ρc = ρ·C b (C b measuring systematic bias).
Lin's concordance and Pearson's correlation values were very similar, suggesting no systematic bias between software packages with an excellent precision ρ for MBF (ρ = 0.97, ρc = 0.96, C b = 0.99) and good precision for MFR (ρ = 0.83, ρc = 0.76, C b = 0.92). On a per-segment basis, no mean bias was observed on Bland-Altman plots, although PMOD provided slightly higher values than FlowQuant at higher MBF and MFR values (P < .0001).
Concordance between software packages was excellent for MBF and MFR, despite higher values by PMOD at higher MBF values. Both software packages can be used interchangeably for quantification in daily practice of Rb-82 cardiac PET.

Keywords
Algorithms, Blood Flow Velocity, Coronary Artery Disease/diagnostic imaging, Coronary Artery Disease/physiopathology, Coronary Circulation, Female, Humans, Image Enhancement/methods, Image Interpretation, Computer-Assisted/methods, Machine Learning, Male, Middle Aged, Myocardial Perfusion Imaging/methods, Pattern Recognition, Automated/methods, Positron Emission Tomography Computed Tomography/methods, Radiopharmaceuticals, Reproducibility of Results, Rubidium Radioisotopes, Sensitivity and Specificity, Software, PET, accuracy, agreement, comparison, concordance, myocardial perfusion, precision, quantification, rubidium-82
Pubmed
Web of science
Open Access
Yes
Create date
17/06/2015 14:11
Last modification date
20/08/2019 12:25
Usage data