Assessment of ECG-gated myocardial SPECT analysis program with cardiac phantom and clinical data

In the phantom study, the effects of ventricular volume, location, absorption, acquisition time, enlarged acquisition and pre-filter on the calculated indices were evaluated. Furthermore, in the clinical study using 99mTc-MIBI, reproducibility between 2 observers, comparison with left ventriculography and effects of pre-filter were assessed. In clinical cases of 201Tl and 123I-BMIPP, left ventricular volume and EF were also analyzed by QGS with various pre-filters. Although the true phantom volumes and calculated volumes showed an excellent linear correlation, calculated volumes were significantly under-estimated by 14.5-33.8%. An absorbent material around the phantom caused reduction in calculated volumes by 4.1-9.1%. Duration of acquisition times, 3 to 60 seconds per projection, did not influence the calculation of the parameters. With enlarged data collection, calculated volume (37 ml) was larger than that of normal acquisition (33 ml). When the cut-off frequency of Butterworth filter was changed, these indices of volume and ejection fraction (EF) were almost stable over 0.41 cycle/cm. There was an excellent correlation in intra-observer measurements for end-diastolic volume (EDV), end-systolic volume (ESV) and EF. In comparison with left ventriculography, correlation of parameters was good in ESV, EF, but was fair in EDV. The QGS program underestimated EDV, ESV and EF. (K.H.)