Measuring perfusion defects in SPECT MPI: A quantitative comparison

Three major quantitative softwares offer cardiologists a comprehensive map of myocardial perfusion and function, but they do not always agree, according to a study in the August issue of the Journal of Nuclear Cardiology.

Sameer Ather, MD, PhD, from the division of cardiovascular disease at the University of Alabama at Birmingham, and colleagues evaluated three commercially available softwares—Quantitative Perfusion SPECT (QPS), Emory Cardiac Toolbox and 4 Dimension-Myocardial SPECT (4DM)—to evaluate the agreement and differences between them in determining perfusion defects and left ventricular (LV) function. Results were a mixed bag.

“Several studies have demonstrated that the above mentioned softwares have a strong correlation with each other, and with an independent ‘gold standard’ in estimating LV function and volumes,” wrote the authors. “In contrast, there are limited data on the degree of agreement between these programs in assessing perfusion abnormalities, and no data regarding agreement of the three softwares in measuring perfusion defect size (PDS) and the size of reversible perfusion defects using polar maps.”

The researchers evaluated 120 consecutive cardiac patients with abnormal regadenoson SPECT MPI with a visual scoring system, and determined agreement between softwares for PDS, and reversible and fixed defect. Results showed poor on up to fair correlations between them. In 62 percent of cases, researchers found significant discordance in defining whether defects were small, moderate or large. However correlation between the softwares improved when evaluating parameters for LV ejection fraction, volumes and mass. In 28 percent of patients it was difficult to get a handle on whether LV systolic dysfunction was normal, mild or severe.

“To the best of our knowledge, this is the first study that has compared the performance of the three softwares in assessing PDS and reversible perfusion defects,” the researchers wrote. “There are several possible reasons for the differences in results obtained with the three software packages.”

These reasons include different algorithms applied to compose a semi-quantitative polar map, which creates a standardized 2D image of 3D LV myocardium in order to compare rest and stress and to use for patient follow-up. Apical sampling was the most striking difference in myocardial sampling between the softwares, whereas methods were better matched in basal and mid-ventricular walls. 

“These results have important implications regarding patient care, as accurate assessment of PDS is vital in guiding clinical decision making,” the authors noted. It may be necessary to use the same software with a given patient to provide more consistency across imaging studies.