[Comparison of left ventricular function evaluation using gated myocardial perfusion imaging with different software...

Area of Science:

• Nuclear Cardiology
• Cardiovascular Imaging
• Medical Physics

Background:

• Accurate assessment of left ventricular (LV) volumes and ejection fraction (EF) is crucial for diagnosing and managing cardiovascular diseases.
• Gated myocardial perfusion imaging (G-MPI) is a widely used non-invasive technique, but its accuracy in quantifying LV function depends on the processing algorithm.
• Comparing different processing algorithms like 4-dimensional model single photon emission computed tomography (4D-MSPECT) and quantitative gated single photon emission computed tomography (QGS) is essential for clinical application.

Purpose of the Study:

• To evaluate the accuracy of 4D-MSPECT and QGS in assessing left ventricular end-diastolic volume (EDV), end-systolic volume (ESV), and ejection fraction (EF) using gated myocardial perfusion imaging (G-MPI).
• To compare the performance of 4D-MSPECT and QGS against left ventriculography (LVG) as the reference standard.
• To determine if algorithm-specific normal databases are necessary for accurate interpretation of G-MPI results.

Main Methods:

• A retrospective analysis of 85 patients who underwent rest G-MPI and LVG within 30 days was performed.
• Left ventricular EDV, ESV, and EF were calculated from G-MPI using both 4D-MSPECT and QGS algorithms.
• Statistical analysis, including correlation coefficients and comparison of means, was used to compare G-MPI results with LVG.

Main Results:

• Both 4D-MSPECT and QGS demonstrated high correlations with LVG for EDV (r=0.89, r=0.81), ESV (r=0.97, R=0.95), and EF (r=0.95, r=0.93).
• Both algorithms significantly underestimated EDV compared to LVG.
• ESV values from both algorithms did not differ significantly from LVG. QGS yielded significantly lower LVEF values than LVG.

Conclusions:

• 4D-MSPECT and QGS show good agreement with LVG for assessing LV EDV, ESV, and EF from G-MPI.
• Both algorithms exhibit slight over- or under-estimations of volumes, indicating potential algorithm-inherent biases.
• The findings suggest the necessity of establishing separate normal databases for each algorithm to ensure accurate clinical interpretation of G-MPI derived LV function parameters.