Dynamic arrhythmia filtration for gated blood-pool imaging: validation against list mode technique.

Normal resting heart rate variations distort the diastolic portions of left ventricular time-activity curves, altering calculated measures of diastolic function such as peak filling rate. Diastolic filling parameters obtained by two methods of arrhythmia removal--list-mode acquisition and a new approach, dynamic arrhythmia filtration (DAF)--were compared. In DAF, data are evaluated for cycle length in real time and accepted or rejected immediately according to preset, operator-determined cycle-length criteria, eliminating the need for postprocessing of data and for large mass data storage. We prospectively determined EF, time to end-systole, peak filling rate, and time to peak filling on gated blood-pool studies of 25 patients. Identical camera and ECG data were sent simultaneously to two computers and analyzed using list mode and DAF, with an R-R interval of 5-6% of mean heart rate. Excellent correlation between list mode and DAF was seen in peak filling rate (r = 0.94), EF (r = 0.99), and time to end-systole (r = 0.97). Lesser correlation seen in time to peak filling (r = 0.65) may be due to inherent problems in measuring this parameter reliably, possibly lessening this parameter's clinical usefulness. DAF is less time consuming and less technically demanding than list mode, and provides results which correlate closely to those obtained by list mode.