Accuracy of left ventricular (LV) volume calculated by Quantitative Gated SPECT (QGS) was studied by using a cardiac phantom with 12 SPECT systems of 10 hospitals. The data were obtained by the same preset time and same preset counts protocols, and also by each hospital's own acquisition protocol. The LV volume calculated by QGS was generally underestimated. The mean errors of the LV volume with each acquisition protocol were 4% in preset time, 5 to 8% in preset counts and greater than 9% in each hospital's own acquisition protocol. To decrease the differences of estimated LV volume among hospitals a uniform protocol for data acquisition should be established.
Uniformity of different types of beta-ray plane sources has been examined. Four types of plane sources, polymer, anodized aluminum, filter and ion exchange membrane, were selected. On estimation of the uniformity, the coefficient of variation with respect to surface emission rate was measured using an imaging plate or a proportional counter. As a result, polymer sources and ion exchange membrane sources showed good uniformity. The area and number of region of interest for estimation of uniformity were discussed.