High frame rate cardiac cine MRI for the evaluation of diastolic function and its direct correlation with echocardiography

BACKGROUND: Breath-hold cine MR is the method of choice for evaluating left ventricular (LV) systolic function; however, the evaluation of diastolic function remains in the domain of high frame rate echocardiography. Thus, a cine MR technique for simultaneously evaluating LV systolic and diastolic function would be clinically valuable. PURPOSE: To test the feasibility of extracting indices that characterize LV diastolic function from high frame rate cine MR. STUDY TYPE: Single center, prospective. POPULATION: Asymptomatic volunteers (N = 24; age 45.8 ± 12.3 years). FIELD STRENGTH/SEQUENCE: High frame rate (70 fps) cine MR and phase-contrast MR during free breathing were acquired at 1.5T. ASSESSMENT: The following MR-based LV filling metrics were extracted from LV volume changes during the cardiac cycle: 1) the volume-rate ratio, REFP /RLFP (ratio of the peak LV filling rate during the early filling period [EFP] to that during the late filling period [LFP]); and 2) the volume ratio, VEFP /VLFP (the ratio of cumulative LV volume change between the EFP and LFP). These metrics were then compared with traditional transmitral blood flow-based MR and echocardiographic indices. The effect of temporal resolution on these metrics was also evaluated. STATISTICAL TESTS: Bland-Altman and linear regression analyses were used to evaluate the performance of the proposed metrics against traditional indices of diastolic function. RESULTS: The REFP /RLFP and VEFP /VLFP correlated well with E/AQ-flow (r 2 = 0.66 and 0.54, respectively) and E/Aecho (r 2 = 0.58 and 0.49, respectively). Systolic indices remained robust ( 8% error) for frame rates below 35 fps. DATA CONCLUSION: In asymptomatic volunteers, cardiac cine MR images acquired at frame rates >35 fps can be used to extract LV diastolic function indices from the temporal changes in LV volume. LEVEL OF EVIDENCE: 1 Technical Efficacy Stage: 1 J. Magn. Reson. Imaging 2019;50:1571-1582.