1356819 - Exploring Heart Rate Variability Differences in Real-Time Compressed Sensing Motion Corrected Cine Imaging in Cardiac MRI

Background: ECG-gated steady-state free precession cine (SSFP) cardiac MRI (CMR) is the gold standard for volumetric analyses of the left (LV) and right ventricles (RV). SSFP image quality can be limited in patients with poor breath-holding ability or arrhythmia, and generally require 5-10 minutes of scanner time between the breath-held segmented acquisition and patient recovery periods. Real-time, free-breathing compressed sensing accelerated, motion-corrected imaging (CS MOCO) is a novel cine SSFP protocol that can decrease scan times and obviates the need for breath-holding. We compare LV and RV volumetric quantification between standard SSFP and CS MOCO.

Methods: Thirty-five patients undergoing CMR were prospectively recruited (61.38 ± 11.50 years; 74% male). Each patient underwent a complete short-axis stack cine of the ventricles (~10 slices) using both SSFP Cine and CS MOCO sequences. Imaging data was analyzed by the same observer (CVI42, Calgary) to attain volumetric data including end-diastolic volume (EDV), end-systolic volume (ESV), cardiac output (CO), and ejection fraction (EF). The cohort was separated in two groups based on RR-interval derived heart rate variability (HRV). HRV was assessed by calculating a coefficient of variation (CoV) using the average RR interval, RR standard deviation, and heart rate. A CoV greater than 0.1 was the threshold for patients with high HRV. The cohorts consisted of 16 high-HRV patients, 17 low-HRV patients. Two patients were excluded, as their RR interval data was inaccessible. Paired t-tests, Bland-Altman analyses, and Pearson coefficient correlations were performed to compare parameters in each cohort.

Results: The mean CS MOCO scan time (2:05 ± 0:35 min) was decreased in comparison to the mean SSFP scan time (4:54 ± 1:16 min). There was also no statistically significant difference in any volumetric or functional parameters (Fig. 1) between the SSFP and CS MOCO sequences in the total cohort. There were moderate to strong positive correlations for each volumetric and functional variable (EDV, ESV, CO, and EF). After categorizing the two cohorts based on HRV, there was no statistically significant difference in any volumetric or functional parameter estimated by the CS MOCO and SSFP imaging sequences for both cohorts (Fig. 1). There were moderate-to-strong positive correlations for the volumetric and functional estimations in the high-HRV cohort; however, there were stronger correlations in the low-HRV cohort. Bland Altman analysis indicates agreement between both protocols, SSFP and CS MOCO, more so in patients with low-HRV (Fig. 2).

Conclusion: CS MOCO sequences provide similar LV and RV volumetric quantification at significantly reduced scan times compared to conventional ECG-gated SSFP. There is an increased agreement in CS MOCO and SSFP scans in patients with low HRV. CS MOCO cine shows promise for improving scan efficiency, quality, and patient experience in CMR clinical practice.