Feasibility of one breath-hold cardiovascular magnetic resonance compressed sensing cine for left ventricular strain analysis

Xiaorong Chen; Jiangfeng Pan; Yi Hu; Hongjie Hu; Yonghao Pan

doi:10.3389/fcvm.2022.903203

Feasibility of one breath-hold cardiovascular magnetic resonance compressed sensing cine for left ventricular strain analysis

Front Cardiovasc Med. 2022 Aug 12:9:903203. doi: 10.3389/fcvm.2022.903203. eCollection 2022.

Authors

Xiaorong Chen¹, Jiangfeng Pan¹, Yi Hu¹, Hongjie Hu², Yonghao Pan¹

Affiliations

¹ Affiliated Jinhua Hospital, Zhejiang University School of Medicine, Jinhua, China.
² Sir Run Run Shaw Hospital, Hangzhou, China.

Abstract

Objective: To investigate the feasibility of 3D left ventricular global and regional strain by using one breath-hold (BH) compressed sensing cine (CSC) protocol and determine the agreement between CSC and conventional cine (CC) protocols.

Methods: A total of 30 volunteers were enrolled in this study. Cardiovascular magnetic resonance (CMR) images were acquired using a 1.436 T magnetic resonance imaging (MRI) system. The CSC protocols included one BH CSC and the shortest BH CSC protocols with different parameters and were only performed in short-axis (SA) view following CC protocols. Left ventricular (LV) end-diastole volume (EDV), end-systole volume (ESV), stroke volume (SV), and ejection fraction (EF) global and regional strain were calculated by CC, one BH CSC, and shortest BH CSC protocols. The intraclass correlation coefficient (ICC) and coefficient of variance (CV) of these parameters were used to determine the agreement between different acquisitions.

Results: The agreement of all volumetric variables and EF between the CC protocol and one BH CSC protocol was excellent (ICC > 0.9). EDV, ESV, and SV between CC and shortest BH CSC protocols also had a remarkable coherence (ICC > 0.9). The agreement of 3D LV global strain assessment between CC protocol and one BH CSC protocol was good (ICC > 0.8). Most CVs of variables were also good (CV < 15%). ICCs of all variables were lower than 0.8. CVs of all parameters were higher than 15% except global longitudinal strain (GLS) between CC and shortest BH CSC protocols. The agreement of regional strain between CC and BH CSC protocols was heterogeneous (-0.2 < ICC < 0.7). Many variables of CVs were poor.

Conclusion: Notably, one BH CSC protocol can be used for 3D global strain analysis, along with a good correlation with the CC protocol. The regional strain should continue to be computed by the CC protocol due to poor agreement and a remarkable variation between the protocols. The shortest BH CSC protocol was insufficient to replace the CC protocol for 3D global and regional strain.

Keywords: agreement assessment; cardiovascular magnetic resonance; compressed sensing; myocardial strain; ventricular function.