Background: Recently faster cardiac magnetic resonance (CMR) cine sequences basing on k-t compressed sensing have been developed.
Purpose: To compare two compressed sensing CMR sequences-one in breath-hold technique and one during free breathing - with the standard SSFP sequence with respect to regional left ventricular function assessment.
Material and Methods: Left ventricular short-axis stacks of two compressed sensing sequences in breath-hold technique (sparse_HB) and during free breathing (sparse_FB; both spatial resolution, 1.8 × 1.8 × 8 mm3) and a standard SSFP cine sequence (spatial resolution, 1.9 × 1.9 × 8 mm3) were acquired in 50 patients on a 1.5 T MR system. Regional wall motion abnormalities (RWMA) were rated qualitatively (normal/hypo-/a-/dyskinesia) by two experienced readers in consensus for all cardiac segments (American Heart Association’s segment model) and sequences. RWMA detection rates were compared between sequences by kappa statistic.
Results: In 13 patients, RWMA were detected in at least one cardiac segment. The RWMA detection rates were similar between CMR sequences (hypokinesia, 7.2% to 7.9%; akinesia, 0.8% to 1.3%; dyskinesia 0.3% to 0.4%) and kappa statistics revealed an almost perfect agreement in RWMA detection between both sparse and the standard SSFP sequence (standard versus sparse_HB: kappa, 0.918, p value, <0.001; standard versus sparse_FB: kappa, 0.868, p value, <0.001).
Conclusion: Compressed sensing cine CMR acquired during breath-hold or free-breathing allows reliable RWMA detection, thus, might alternatively be used in cine CMR for regional left ventricular function assessment.