Single breath-hold magnetic resonance cine imaging for fast assessment of global and regional left ventricular function in clinical routine
We aimed to investigate comparability of LV volumes, function, and mass acquired with three steady-state free precession (SSFP) pulse sequences, simulating typical vendor and protocol specific differences in data acquisition. Twenty-one healthy subjects (11 male and 10 female; age range 23–49) underwent cardiac magnetic resonance (CMR) imaging at 1.5 Tesla (T). A complete stack of short-axis views covering the entire left ventricle (LV) were acquired for each of the three SSFP sequences, differing in the interslice gap and slice thickness (7 mm with no gap (7/0 mm); 7 mm with a 3 mm gap (7/3 mm) and 6 mm with a 4 mm gap (6/4 mm)) with slight variations in acquisition parameters. For each sequence, the LV volumes, function, and mass were determined. Intra- and inter-observer variability and inter-study reproducibility were assessed for all protocols. All LV volumes, function and mass parameters were similar for the three SSFP sequences (P > 0.05 for all). The LV ejection fraction for the 7/3 mm sequence was 67.2 ± 6.0, 67.4 ± 5.3 for the 7/0 mm sequence, and the 6/4 mm sequence was 69.2 ± 5.7. The LV mass ranged from 119.8 ± 32.4 for the 7/3 mm sequence to 122.2 ± 34.0 for the 7/0 mm sequence. Variabilities were low with no difference in variability between the sequences. The three SSFP pulse sequence techniques resulted in similar LV volume, function, and mass measurements with no difference in observer and interstudy variabilities. This may allow application and transfer of LV volume studies and databases based on different imaging parameters, at different CMR sites, with a given post-processing method. Future multi-centre studies may now be in a position to consider multi-vendor study designs for LV volume studies.