Accelerated 4D dobutamine stress MR imaging with k-t BLAST: feasibility and diagnostic performance.

PURPOSE: To prospectively determine feasibility and diagnostic performance (with angiography as reference standard) of k-space and time (k-t) broad-use linear acquisition speed-up technique (k-t BLAST) cine imaging during dobutamine stress for identification of inducible cardiac wall motion abnormalities.

MATERIALS AND METHODS: The study was conducted according to standards of the Charité and Virchow-Klinikum Ethics Committee. Patients gave written consent. Dobutamine stress magnetic resonance (MR) imaging was conducted in 65 patients (mean age, 63 years +/- 9 [standard deviation]; 49 men) with conventional cine steady-state free precession (SSFP). Accelerated four-dimensional (4D) k-t BLAST single-breath-hold imaging with complete left ventricular (LV) coverage was also performed at rest and during stress. For the cine SSFP and accelerated cine techniques, duration of imaging at rest and LV end-diastolic volume and ejection fraction were assessed. Segmental agreement for resting and inducible wall motion abnormalities was determined. In a subgroup (n = 40), direct comparison between SSFP and accelerated cine was performed for coronary stenosis detection. A paired Student t test was used to assess significance of continuous variables. Pearson correlation was used to test correlation between the techniques. Sensitivity, specificity, and diagnostic accuracy were calculated (standard definitions). For quantitative measurement of agreement, Cohen kappa was applied.

RESULTS: For accelerated cine, imaging duration at rest was shortened by 40%. Correlations between cine SSFP and accelerated cine for LV parameters were 135 mL +/- 37 versus 129 mL +/- 31 (r = 0.89) for end-diastolic volume and 59% +/- 8 versus 58% +/- 7 (r = 0.95) for ejection fraction. kappa Values for segmental wall motion at rest and stress ranged from 0.77 to 0.91. Sensitivity, specificity, and diagnostic accuracy for coronary stenosis (>or=50%) detection based on arterial territory were 82%, 87%, and 86%, respectively, for cine SSFP and 82%, 86%, and 85%, respectively, for accelerated cine imaging.

CONCLUSION: Accelerated 4D k-t BLAST wall motion imaging at rest and at dobutamine stress is rapid and feasible; LV measurements were nearly identical between the imaging approaches. Segmental wall motion analysis at rest and at stress show excellent agreement and reliable depiction of myocardial territories supplied by coronary arteries with 50% or more luminal narrowing.