Reproducibility of assessment of myocardial function using gated 99Tc(m)-MIBI SPECT and quantitative software

We investigated the reproducibility of the quantification of left ventricular volumes and ejection fraction, and grading of myocardial wall motion and systolic thickening with gated myocardial single-photon emission computed tomography (SPECT) and Cedars quantification software.

We performed post-stress gated myocardial SPECT in 33 consecutive patients twice sequentially in situ. Sixteen frames per cycle were used for gating. Using Cedars quantitative gated SPECT software ventricular volumes and ejection fraction (EF) were calculated. Wall motion and thickening was graded using 5- and 4-point scores, respectively. Coefficients of variation for re-examination of volumes and EF were calculated. Kappa values for assessing reproducibility of wall motion or wall thickening were calculated.

Root mean square of errors was 5.0 ml for end-diastolic volume, 3.9 ml for end-systolic volume and 1.9% for EF. No bias or difference was found between the first and second acquisition by Bland-Altman analysis. Kappa values for wall motion and thickening of repeated acquisition was 0.76 and 0.87, respectively. The value of 2 SD from Bland-Altman analysis was 14.4 ml for EDV, 11.2 ml for ESV and 5.3% for EF.

We conclude that quantification of functional indices and assessment of wall motion or thickening using gated 99Tc(m)-sestamibi (MIBI) SPECT was reproducible and we found that their ranges of physiological fluctuation were narrow enough.