Related Experiment Videos
Left ventricular function parameters obtained from gated myocardial perfusion SPECT imaging: a comparison of two data
Deborah A Tout1, Angela Rogers, Andries Van Aswegen
1Imperial College London, National Heart and Lung Institute, Royal Brompton Campus, London, UK. d.tout@rbh.nthames.nhs.uk
Nuclear Medicine Communications
|January 20, 2005
Summary
Reproducibility of Quantitative Gated SPECT (QGS) for left ventricular function analysis varies by processing system. Differences arise from reconstruction, not analysis, highlighting the need for standardized QGS protocols.
Area of Science:
- Nuclear Cardiology
- Medical Imaging Analysis
- Cardiovascular Diagnostics
Background:
- Quantitative Gated Single Photon Emission Computed Tomography (SPECT) (QGS) is crucial for assessing left ventricular function from myocardial perfusion scintigraphy (MPS).
- Previous validation of QGS exists, but its reproducibility across different processing systems remains under-explored.
- This study investigates QGS performance on workstations from two distinct manufacturers.
Purpose of the Study:
- To compare the reproducibility of left ventricular function parameters calculated by QGS on different processing systems.
- To identify potential discrepancies in QGS results stemming from data reconstruction versus analysis software.
Main Methods:
- Retrospective analysis of 50 gated rest MPS studies.
- Filtered back-projection (FBP) reconstruction using identical parameters on Philips Pegasys and Nuclear Diagnostics Hermes workstations.
- QGS analysis of end-diastolic volume (EDV), end-systolic volume (ESV), and left ventricular ejection fraction (LVEF) on both systems, including cross-system analysis of reconstructed data.
Main Results:
- Significant mean differences in EDV, ESV, and LVEF were observed when QGS data were reconstructed on separate systems (-7.06 ml EDV, -5.54 ml ESV, +1.14% LVEF).
- Bland-Altman plots revealed trends for EDV and ESV differences with separate reconstructions.
- Minimal differences in EDV, ESV, and LVEF were found when the same reconstructed data were analyzed on both systems (-0.16 ml EDV, -0.10 ml ESV, +0.14% LVEF), with no significant trends.
Conclusions:
- Discrepancies in QGS-derived left ventricular function parameters between different systems are primarily due to variations in the image reconstruction process.
- Identical QGS analysis of the same reconstructed data yields highly reproducible results across different workstations.
- Standardization of reconstruction protocols is essential for ensuring consistent QGS quantification across diverse processing platforms.