High frame rate cardiac cine MRI for the evaluation of diastolic function and its direct correlation with echocardiography. Issue 5 (21st May 2019)

• Record Type: Journal Article
• Authors: Zhang, Jiming; Chen, Jie; Cheong, Benjamin; Pednekar, Amol; Muthupillai, Raja
• Is Part Of: Journal of magnetic resonance imaging. Volume 50:Issue 5(2019)
• Page Start: 1571
• Page End: 1582
• Publication Date: 2019-05-21
• Subjects: magnetic resonance imaging -- SSFP -- echocardiography -- diastolic function -- systolic function -- phase contrast; Magnetic resonance imaging -- Periodicals; 616
• DOI: 10.1002/jmri.26791
• Languages: English
• ISSNs: 1053-1807

Abstract

Abstract : 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 (<3% error) for frame rates ≥20 fps. Although the proposed VEFP /VLFP was robust (<5% error) up to 25 fps, the proposed volume–rate diastolic function metrics were less reliable (>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.