Myocardial perfusion and left ventricular quantitative parameters obtained using gated myocardial SPECT: Comparison of three software packages

• Published in: Journal of nuclear cardiology 2018-06, Vol.25 (3), p.911-924
• Main Authors: Alexiou, Sotiria, Georgoulias, Panagiotis, Angelidis, George, Valotassiou, Varvara, Tsougos, Ioannis, Psimadas, Dimitrios, Lakiotis, Velissarios, Kaspiri, Agaristi, Alexopoulos, Dimitrios, Apostolopoulos, Dimitrios, Vassilakos, Pavlos
• Format: Article
• Language: English
• Subjects: Aged; Algorithms; Cardiology; ECTb; Ejection fraction; Exercise Test; Female; Gated Blood-Pool Imaging; Gated SPECT; Humans; Imaging; left ventricular ejection fraction; left ventricular volumes; Male; Medicine; Medicine & Public Health; Middle Aged; Myovation; Nuclear Medicine; Original Article; QGS; Radiology; Reproducibility of Results; Sensitivity and Specificity; Software; Software packages; Stroke Volume - physiology; Technetium; Tomography, Emission-Computed, Single-Photon; Ventricular Dysfunction, Left - diagnostic imaging; Ventricular Dysfunction, Left - physiopathology
• ISSN: 1071-3581; 1532-6551
• DOI: 10.1007/s12350-016-0730-x

The aim of the present study was to compare Emory Cardiac Toolbox, Myovation, and Quantitative Gated SPECT software regarding the automatic measurements of perfusion and functional left ventricular (LV) quantitative parameters, summed stress score (SSS), perfusion defect score, LV ejection fraction (LVEF), end-diastolic volume, and end-systolic volume (ESV). 99mTc-tetrofosmin gated SPECT studies were performed in 634 consecutive patients based on the one-day stress/rest protocol. Participants were divided into subgroups according to heart size (ESV cut-off value: 25 mL), perfusion (SSS >/≤3), and other patient/protocol-related factors. LVEF was categorized as normal (≥50%), mildly moderately impaired (35–49%), and severely abnormal (25 mL, and SSS >3 subgroups (intraclass correlation coefficients, ICCs 0.73–0.93). In SSS ≤3 subgroup, the correlation was excellent for LV functional parameters, but suboptimal for perfusion variables (ICCs 0.30–0.83). LVEF categorization revealed similar variability (discordance 18.1 and 11.1% for stress/rest LVEF values, respectively). Pair comparisons demonstrated considerable differences concerning all parameters for all patient subgroups. The statistical significance of our findings by ESV and SSS classifications was evaluated. Despite the significant concordance between software packages, considerable differences in mean values of myocardial perfusion and LV functional parameters were demonstrated.