Abstract
Background: Single Photon Emission Computed Tomography (SPECT) myocardial perfusion imaging (MPI) is a non-invasive method used to detect myocardial ischemia, which helps assess left ventricular function when used in conjunction with electrocardiogram gating. While quantitative analysis software greatly improves diagnostic precision in cardiology, it can still be affected by variations in operator technique.
Purpose: To assess the reproducibility of cardiac function parameters in gated SPECT MPI.
Methods: Two nuclear medicine doctors with over 3 years of experience in SPECT/CT image reading independently utilized 4DM software to measure left ventricular function parameters with the ROI delineation method recommended by the manufacturer. Initial measurements were conducted, followed by a re-measurement after one week. Measurement results obtained by the same physician during both sessions were categorized as ‘within-investigator’ consistency, while those acquired by different physicians were designated as ‘between-investigators’ consistency.
Intraclass correlation coefficients (ICCs) were employed to evaluate consistency. Furthermore, Bland-Altman analysis was conducted to examine the agreement in discrepancies among different indicators. Simultaneously, the least significant changes (LSCs) and reproducibility of gated SPECT MPI were calculated and measured.
Results: The intra-observer ICC values of left ventricular ejection fraction (LVEF), end-diastolic volume (EDV), end-systolic volume (ESV), and summed stress score (SSS)/summed rest score (SRS) were 0.973-0.978, 0.995-0.997, 0.983-0.988, and 0.976-0.991, respectively. The inter-observer ICC values were 0.954-0.973, 0.990-0.997, 0.973-0.991, and 0.976-0.992, respectively. Gated SPECT MPI measured by Observer 1 and Observer 2 showed very high consistency in measuring LVEF, EDV, ESV, and SSS/SRS. The results of reproducibility assessment showed that the root-mean-square standard deviation (SDRMS) of LVEF, EDV, ESV, and SSS/SRS were 1.86-2.20%, 1.16-1.27%, 0.94-1.17%, and 0.56-0.92%, respectively. The corresponding LSCs were 5.15-6.10%, 3.22-3.53%, 2.62-3.16%, and 1.54-2.54%, respectively.
Conclusions: The region of interest (ROI) delineation method recommended by the device manufacturer was adopted and employed uniformly in the medical team to benefit from the repeatability measurement of cardiac function parameters in gated SPECT MPI. In addition, basic tests such as LSCs and reproducibility can be included in the quality assurance plan, which can be conducted regularly. Consequently, every medical department can establish its control framework through these fundamental experiments.