Gated cardiac blood pool studies in atrial fibrillation: role of cycle length windowing

Cycle length windowing is gaining increasing acceptance in gated blood pool imaging of patients with atrial fibrillation (AF). The goals of this study were: to assess differences of ejection fraction (EF) in AF with and without windowing and to determine how EF varied with cycle length in patients with AF. Twenty patients with AF were prospectively studied by gated blood pool imaging, with simultaneous collection in each patient of 5-7 studies with cycle length windows spanning the cycle length histogram. Each window accepted beats of only a narrow range of cycle lengths. EF was determined for each of the narrow cycle length windows as well as for the entire gated blood pool study without cycle length windowing. For every patient an average of the windowed EFs was compared with the non-windowed EF. EF values were similar (mean windowed: 46.6; non-windowed: 45.5; P=0.16), and there was a good correlation between the two techniques (r=0.97). The data were then examined for a relationship of EF with cycle length. The difference from average windowed EF (ΔEF) was calculated for each window and plotted vs. the cycle length of the center of each window. No predictable linear or nonlinear relationship of ΔEF with window position was observed. Lack of predictable variation of EF with cycle length is likely due to lack of a predictable amount of ventricular filling for a given cycle length, as the amount of diastolic filling in AF depends on the random cycle length of the preceding beat. In summary, windowing in AF does not provide a clinically significant difference in EF determination. If cycle length windowing is used, the exact location of the window is not critical.