Assessment of left ventricular systolic function is an important clinical variable with respect to diagnosis, prognosis and treatment in various clinical situation. Automatic quantification of ventricular volume and ejection fraction by delineating 3 dimensional endocardial boundaries from the gated perfusion SPECT was validated. The purposes of this study were to assess the agreement of left ventricular ejection fraction determined by two-dimensional echocardiography and automatic quantification of perfusion SPECT and compare different echocardiographic methods with the reference method, automatic quantification of perfusion SPECT.

Twenty seven patients (20 men, 7 women; mean age 60±12) with acute myocardial infarction (anterior: 16, inferior: 7, lateral: 4) and twenty one patients (12 men, 9 women; mean age 60±12) without myocardial infarction history and regional wall motion abnormality in echocardiography were studied. All patients had two-dimensional echocardiography and 201Tl perfusion SPECT performed within 5 days of each other. Ejection fraction of left ventricle was calculated by echocardiography using modified Simpson's method and M-mode method. Also, ejection fraction of left ventricle was estimated by 201Tl perfusion SPECT using automatic software (quantitative gated SPECT[QGS]; Cedars-Sinai Medical Center, Los Angels, CA).

The agreement of ejection fraction between M-mode method and QGS (limit of agreement -3.48, 3.2; average limit of agreement 6.68) in control group was better than that between modified Simpson's method and QGS (limit of agreement 0.04, 8.42; limit of agreement average limit of agreement 8.38). But, The agreement of ejection fraction between modified Simpson's method and QGS in the acute myocardial infarction group (limit of agreement; -15.31, 10.01; average limit of agreement 12.66) was better than that between M-mode method and QGS (limit of agreement -17.82, 13.86; average limit of agreement 15.84). The agreement of ejection fraction between modified Simpson's method and QGS (limit of agreement; -8.49, 5.74; average limit of agreement 7.12) in the anterior myocardial infarction was more accordancy than lateral and inferior wall infarction (limit of agreement; -12.11, 13.74; average limit of agreement 12.92).

The M-mode method in patients without regional wall motion abnormality and modified Simpson's method in patients with regional wall motion abnormality, especially in anterior wall infarction seen to be useful method for the assessment of left ventricular ejection fraction.