AUTOMATION OF GATED TOMOGRAPHIC LEFT-VENTRICULAR EJECTION FRACTION

Background The feasibility of determining left ventricular (LV) ejecti on fraction (EF) from Tc-99m-labeled sestamibi gated tomography (GSPEC T) is well established, To improve precision of measurement, rules use d by observers in processing tomograms were encoded for automation. Me thods and Results, LV centers were estimated from activity centroids o f time-difference images exceeding 50% of maximum counts, End diastole and end systole were defined by time-varying maximum count extremes, Endocardial borders were generated by fitting maximum locations with f ifth-order two-dimensional harmonics, searching inward to predetermine d thresholds, and reconciling endocardial with valve plane points, Reg ression analysis of GSPECT EF yielded r = 0.87 versus equilibrium gate d blood pool in 75 patients and r = 0.87 versus gated first pass in 65 patients, GSPECT EF interobserver variability was r = 0.92 and intrao bserver automatic versus manual linear correlation was r = 0.94. A sub group of 25 studies was analyzed by six independent observers, for who m EF agreement with the core laboratory ranged from r = 0.93 to r = 0. 96, Experienced observers judged it necessary to alter end-diastolic o r end-systolic frames in 7% of patients, endocardial borders in 14%, a nd LV centers in 28%. Conclusion. Results of automated GSPECT LV EF co rrelated well with those of manual GSPECT and gated first-pass and equ ilibrium blood pool values and were highly reproducible.