Contrast-enhanced harmonic color Doppler for left ventricular opacification: Improved endocardial border definition compared to tissue harmonic Imaging and optimization of methodology in patients with suboptimal echocardiograms

Objectives: This study compared the efficacy of contrast-enhanced harmonic color Doppler (C-HCD) and tissue harmonic imaging (THI) for left ventricular endocardial border delineation and explored the optimal methodology of C-HCD in patients with suboptimal echocardiograms. Background: The value of C-HCD in improving endocardium remains unknown. Effects of harmonic velocity-encoded color Doppler (HVD) and harmonic power Doppler (HPD) as well as contrast administration and image acquisition modalities on left ventricular opacification (LVO) have not been established. Methods: One hundred (50 HVD, 50 HPD) patients with suboptimal echocardiograms during conventional fundamental echocardiography were studied with THI and C-HCD using Levovist. Each patient underwent different random contrast administration and image acquisition modalities. Endocardial border definition score index (EBDI), blooming artifacts, contrast destruction, and salvage of suboptimal echocardiograms were calculated in each patient after contrast enhancement. Results: EBDI improved from 2.05 +/- 0.61 in THI to 2.73 +/- 0.48 in HVD, and 1.98 +/- 0.73 in THI to 2.69 +/- 0.51 in HPD (both P < 0.001). The conversion of a nondiagnostic image from fundamental echocardiography to an optimal diagnostic image was 33 (33,70) patients in THI compared to 77 (77%) patients in C-HCD (P < 0.001). Blooming artifacts were seen more commonly in HVD than HPD, intermittent than continuous image acquisition, and bolus than infusion administration (all P < 0.001). There was less contrast destruction in intermittent compared with continuous image acquisition (P < 0.001). Contrast destruction was similar in HVD and HPD, bolus and infusion injection of contrast. The highest salvage rate of a nondiagnostic image from THI to an optimal diagnostic image was 45.5% and 42.4% in HPD mode, with intermittent image acquisition during bolus and infusion contrast administrations. Conclusions: C-HCD seems more effective in demonstrating improved endocardial border definition compared to THI. HPD has less blooming artifacts compared with HCD. The optimal method for LVO was to use HPD with intermittent image acquisition during bolus or infusion administration of Levovist.