We present a simultaneous gated SPECT (G-SPECT) dual-isotope technique using Tl-201 for perfusion and Tc-99m blood-pool labeling for function imaging. Methods: Seventeen patients (13 with previous myocardial infarction, MI) and a control group of three normal volunteers were investigated. They received, 15 min after a Tl-201 stress/redistribution protocol with reinjection, 900-950 MBq Tc-99m-HSA for blood-pool labeling. Eight frames per R-R interval were recorded in the G-SPECT mode with three windows: window A with 20% centered at 71 keV for Tl-201, window B with 10% centered at 105 keV for Te scatter contamination and window C centered at 140 keV with 20% for Tc-99m. Nongated, crosstalk-corrected Tl-201 SPECT perfusion images were reconstructed according to normalized projection-by-projection subtraction from data from windows A and B. G-SPECT data from window C were reconstructed with the same reconstruction limits to allow topographic correlations of left ventricular perfusion and wall motion abnormalities. Polar maps of perfusion and function were used to divide the myocardium into 20 segments. Perfusion was expressed as the percentage of thallium uptake and function corresponding to diastolic to systolic shortening normalized by end diastolic volume, Results: Segmental comparison of uncontaminated-to-contaminated and corrected Tl-201 patient images demonstrated an overall agreement score of 93%, with a kappa statistic of 0.76 +/- 0.06 when normal perfused segments were excluded. Segmental matching of perfusion against function at rest showed no correlation far the 10 patients with preserved ejection fraction of 59% +/- 8% nor for the control group, For the remaining seven patients with an ejection fraction of 34% +/- 10%, there was linear correlation between perfusion and function (r(2) = 0.61). Conclusions: The feasibility of dual Tl-Tc G-SPECT was examined at rest and suggests low perfusion hypokinesis that matches linear dependence for CAD patients with low ejection fraction.
