Multi-view Three-Dimensional Fusion Echocardiography System: First Pilot Study in Patients

Abstract

Background: Multi-view three-dimensional fusion echocardiography (M3DFE) has been shown to improve image quality compared to standard three-dimensional echocardiography (3DE). The clinical application however has been limited due to the fact that suitable recordings require longer breath holds and very stable patient positioning which is not applicable to many patients. To address these limitations, an advanced Three-Dimensional (3D) fusion system has been developed. This thesis examines the hypothesis that this advanced M3DFE system can be successfully applied for the measurement of left ventricular (LV) function in patients with heart failure treated with cardiac resynchronization therapy (CRT) devices. Methods: Patients with heart failure treated with CRT devices were approached for study enrollment during their standard two-dimensional contrast echocardiography (2DCE) visit. The M3DFE protocol was applied after the standard 2DCE protocol and consisted of 3 phases: recording, alignment and fusion. Participants’ datasets were classified into three different groups: single-view three-dimensional echocardiography (S3DE), M3DFE, and standard 2DCE. The percentage of participants undergoing successfully each phase of the protocol was evaluated. Visual LV endocardial border definition (EBD) of M3DFE and S3DE datasets was graded by 2 independent readers. Each reader’s M3DFE EBD score was compared to S3DE EBD score. The global and regional LV systolic function of M3DFE datasets was evaluated by 2 readers and compared to standard 2DCE. Results: Twelve heart failure patients treated with CRT devices were enrolled in the study. 11/12 (91.7%) participants successfully underwent the recording phase of the M3DFE protocol. 99/108 (91.7%) S3DE datasets could be successfully recorded. The alignment and fusion software could be successfully applied in 96/108 (88.9%) of S3DE datasets. The mean improvements in EBD score by the 2 independent readers in the iii M3DFE group were as follows: 24.0±3.3 (95% CI, 19 and 27) and 24.3±3.8 (95% CI, 17 and 28). The corresponding values in the S3DE group were 11.7 ±6.0 (95% CI, 3 and 24) and 10.5 ±5.6 (95% CI, 2 and 24), p<0.01). The mean and standard deviation of the LV ejection fraction (EF) was 39.5±14.8 by reader 1 in M3DFE group, 37.7±13.1 by reader 2 in M3DFE group, and 40.30±15.7 by the reader of the standard 2DCE. The EF measured on M3DFE datasets was not significantly different from that measured by standard 2DCE (p > 0.05). The percentage of agreement in assessing the degree of consistency among the 2 readers in evaluating the regional LV systolic function was 83.3% between the 2 readers of the M3DFE group, 76.5% between reader 1 of the M3DFE group and the reader of the standard 2DCE group, and 74.2 % between reader 2 of the M3DFE group and the reader of the standard 2DCE group. Conclusion: The M3DFE protocol is feasible in patients with heart failure treated with CRT devices. The LV EBD score in M3DFE datasets is superior to S3DE and the assessment of LV systolic function in M3DFE group is comparable to 2DCE group.