[OA107] Validation of left ventricular ejection fraction with gated IQ-SPECT myocardial perfusion imaging. (August 2018)

• Record Type: Journal Article
• Authors: Leva, Lucia; Matheoud, Roberta; Sacchetti, Gian Mauro; Brambilla, Marco
• Is Part Of: Physica medica. Volume 52(2018)Supplement 1
• Page Start: 41
• Publication Date: 2018-08
• Subjects: Medical physics -- Periodicals; Biophysics -- Periodicals; Biophysics -- Periodicals; Imagerie médicale -- Périodiques; Radiothérapie -- Périodiques; Rayons X -- Sécurité -- Mesures -- Périodiques; Physique -- Périodiques; Médecine -- Périodiques; 610.153
• DOI: 10.1016/j.ejmp.2018.06.179
• Languages: English
• ISSNs: 1120-1797

Abstract

Abstract : Purpose: This study was motivated by the inconsistency reported in myocardial perfusion imaging (MPI) studies both in phantom and in patients in the reported ejection fraction (EF) and perfusion indexes between gamma cameras equipped with confocal collimators (IQ-SPECT) and conventional low-energy high-resolution (LEHR) collimation protocols. The aim of the study was to assess in patients the agreement between the EF values and perfusion indexes obtained with IQ-SPECT and reconstructed with iterative reconstruction algorithms and with a conventional gamma camera equipped with LEHR (CGC) and reconstructed with Filtered Backprojection (FBP), considered as the gold standard. Methods: Gated-stress myocardial perfusion imaging (MPI) was performed in 56 patients with suspected coronary artery disease. The patients underwent in a random order two sequential acquisitions performed on Symbia-IQ SPECT (4.8 mm pixel size, 1.6°/step, 28"/step, 208° angular range, reconstructed with OSEM, 10x3 iterations, 10 mm-Gaussian filter) and on a Ecam SPECT equipped with LEHR collimators (6.6 mm pixel size, 2.6°/step, 25"/step, 180 angular range, reconstructed with FBP, Butterworth filter 8–0.45); a subset of 32 patients underwent in a separate day a Gated-rest MPI, after injection of an average 524 ± 44 MBq of 99m Tc-tetrofosmin. EF values and perfusion indexes (summed stress score-SSS, summed rest score-SRS), were evaluated separately for stress and rest MPI and compared as (CGC-IQ) using Bland–Altman plots estimating the bias and the limits of agreement. Results: For EF, the bias were −4.1% −2.1% and the limits of agreement ranged from −18.0% to 9.9% and from −18.2% to 14.1% for stress and rest MPI, respectively. For SSS, bias and limits of agreement were 2.7 and [−9.9; +15.3] and for SRS they were 2.0 and [−5.8; +9.8]. Conclusions: IQ configuration overestimates EF values for both stress and rest studies with bias of 2–4% and a considerable amplitude in LA interval that spans about 30%. Thus, there is a tendency to overestimate EF with IQ-SPECT. Of still more concern, is that the limits of agreement between the CGC and IQ are wide indicating that the reconstructed datasets are not interchangeable, and the values obtained with IQ-SPECT should only be used with caution when evaluating the functional state of the heart.