Purpose: To evaluate the accuracy of cardiac function measured with real-time true fast imaging with steady-state precession (True-FISP) cine without breath-hold compared with those measured from segmented True-FISP cine with breath-hold.
Methods: Eighteen consecutive patients and six healthy volunteers were enrolled in the study group. Both real-time multislice True-FISP cine imaging without breath-hold and single-slice segmented True-FISP cine imaging with multiple breath-holds were performed in short-axis imaging sections to encompass the entire ventricles. Vertical long-axis cine imaging using real-time True-FISP cine sequence without breath-hold was performed to evaluate heart motion during respiration in 13 subjects. Ventricular volume and mass were evaluated by four observers independently with manual tracing.
Results: Real-time True-FISP cine quality was sufficient for contour detection in all 24 subjects. Cardiodynamic measurements based on real-time True-FISP cine correlated well with those based on segmented True-FISP cine [left ventricular (LV) end-diastolic volume: r = 0.98; LV end-systolic volume: r = 0.98; LV ejection fraction: r = 0.91; LV mass: r = 0.96; right ventricular (RV) end-diastolic volume: r = 0.89; RV end-systolic volume: r = 0.94; RV ejection fraction: r = 0.79]. Intra- and interobserver variability were sufficiently small in real-time True-FISP cine without breath-hold. Heart motion during respiration along the long axis of the left ventricle (2.2 mm to 3.7 mm) was much less than the slice interval (10 mm), confirming that misregistration of slice position during respiration was low.
Conclusion: Real-time True-FISP cine without breath-hold has high reproducibility and is applicable to patients with severe cardiac dysfunction and/or arrhythmias.