Introduction First pass cardiac magnetic resonance (CMR) myocardial perfusion imaging (MPI) requires the rapid acquisition of large amounts of image data. The advanced acceleration technique, k-t SENSE, exploits spatiotemporal correlations to speed up data acquisition. In CMR MPI, k-t SENSE has been used to improve spatial resolution, temporal resolution or slice coverage with reductions in endocardial dark rim artefacts reported. To date, there has been no direct comparison of these strategies.

Method Adenosine stress and rest MPI was performed using a 1.5T Philips Intera system. Ten healthy volunteers were scanned on four occasions using a different saturation recovery gradient echo perfusion sequence at each visit (abstract 079 table 1). The order of sequence utilisation varied for each volunteer.

Image analysis was performed on the middle slice. Image quality was scored 0–3 (0-poor, 3-excellent) and breathing artefacts noted by two readers in consensus. Dark rim artefact was assessed by: thickness (mm), duration (frame count) and extent (area of artefact contoured and expressed as a percentage of myocardial area). Semi-quantitative myocardial perfusion reserve index (MPRI) and Fermi deconvolution derived myocardial perfusion reserve (MPR) were calculated using standard techniques for each sequence type in seven of the volunteers. Friedman Test and repeated measures Analysis of Variance Testing (with Bonferroni correction for pairwise comparisons) were used to compare non-parametric and parametric data respectively.

Results Image quality at stress did not significantly differ between the sequences (abstract 079 table 2).

At rest, the k-t Hybrid sequence showed highest quality. Rim artefact thickness was similar using REFERENCE and k-t Fast sequences but significantly lower using k-t High and k-t Hybrid sequences. Rim artefact extent was significantly higher for REFERENCE images, similar for k-t Fast and k-t Hybrid images and significantly lower for k-t High images. Although four-way comparison suggested differences at rest, rim artefact duration did not differ between sequences by pairwise analysis at stress or rest. MPRI and MPR results did not differ between the four sequences (Mean MPRI:REFERENCE 1.67, k-t High 1.87, k-t Fast 1.94, k-t Hybrid 1.71, p=0.265. Mean MPR:REFERENCE 2.49, k-t High 2.56, k-t Fast 2.53, k-t Hybrid 2.67, p=0.58). Breathing artefacts occurred in 57% of stress and 17% of rest k-t SENSE studies. No breathing artefacts were seen with REFERENCE data.

• Download figure
• Open in new tab
• Download powerpoint

Abstract 079 Figure 1

Conclusion Although spatial and temporal resolution both influence the thickness and extent of dark rim artefact, maximising spatial resolution by k-t SENSE acceleration produces the greatest reduction in these parameters. Semi-quantitative and fully quantitative measures of myocardial perfusion were comparable between k-t SENSE and SENSE acceleration techniques.