TY - JOUR T1 - Gated SPECT processing with QGS: Left ventricular function depends on the tomographic reconstruction method used JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 1133 LP - 1133 VL - 52 IS - supplement 1 AU - Doumit Daou AU - Carlos Coaguila AU - Mark Tawileh Y1 - 2011/05/01 UR - http://jnm.snmjournals.org/content/52/supplement_1/1133.abstract N2 - 1133 Objectives Gated SPECT myocardial perfusion (GSPECT) has been extensively validated for the quantification of left ventricular (LV) function especially with QGS. This was done with filtered backprojection reconstruction method (FBP). Recent developments in nuclear medicine technologies allow simple application of iterative reconstruction with and without resolution recovery. We aimed to compare the performance of the QGS software for the quantification LV function when using FBP as compared to iterative reconstruction (OSEM) and iterative reconstruction combined to resolution recovery (3D-Flash, Siemens). Methods Our study included 51 consecutive patients addressed for myocardial perfusion scintigraphy. Studies were acquired on a two-headed gamma-camera (Symbia, Siemens). GSPECT studies were reconstructed using 3 different methods (FBP, OSEM, and 3D-Flash) and then processed with the QGS software. LV end diastolic volumes (EDV), end systolic volumes (ESV) and LVEF were compared. Results LVEF was higher with FBP (72±13%) than OSEM (70±11%, P<0.0001) and 3D-Flash (69±12%, P<0.0001) respectively. LV EDV was lower with FBP (70±23 ml) than OSEM (74±25 ml, P<0.0001) and 3D-Flash (80±26 ml, P<0.0001) respectively. LV ESV was lower with FBP (22±14 ml) than OSEM (24±16 ml, P<0.0001) and 3D-Flash (2717 ml, P<0.0001) respectively. Bland-Altman analysis for the combined EDV and ESV (n=102) between FBP and 3D-Flash showed that the difference in LV volumes between the 2 methods increased with their average LV volumes (r=0.61, P<0.0001). This also was also verified for FBP and OSEM: r=0.44 (P<0.0001). Conclusions Newer iterative reconstruction methods (OSEM, 3D-Flash) give significantly different LVEF and volumes than those obtained with FBP. This should be considered when changing reconstruction algorithms in clinical practice ER -