Usefulness of the new imaging technique, second harmonic with and without contrast in endocardial border visualization. Reliability analysis in segmentarial wall motion assessment

Objectives. This study sought to determine if newer techniques significantl y improve endocardial border definition in suboptimal acoustic windows, and the reproducibility of the evaluation of wall motion abnormalities accordi ng to the different techniques and degrees of expertise. Methods. We studied a total of 20 consecutive patients with poor ultrasound window, to assess, if the use of tissue harmonic imaging (2H) or contrast with second harmonic (Levovist(R); 4 g i.v.), (2HC) improves endocardial bo rder visualization. In order to analyze inter and intraobserver reliability with the different techniques, four observers with different degrees of ex pertise were each asked to assess the segmental wall motion score of 31 con secutive echocardiograms. Results. The quality of the image was clearly superior with 2H and 2HC comp ared with 2D. This difference was larger in apex and lateral endocardial bo rder from 0.9 and 1 to 1.5 and 1.64 (p < 0.001) with 2H. 2HC was found to s lightly but significantly improve the endocardial definition in apex compar ed with 2H (1.64 vs 1.81; p = 0.016). The percentage of segments assessed for interobserver variability significa ntly improve with 2H and 2HC (2D = 50%, 2H = 75% and 2HC = 95%). Interobser ver agreement with the different techniques between the experienced observe rs did not statistically differ. The less experienced observer presented a significantly lower interobserver reliability than those with experience, a nd did not improve with 2H and 2HC. Conclusions. a) Native tissue harmonic imaging and second harmonic imaging with contrast (Levovist(R)) significantly improves endocardial border visua lization; b) the newer imaging techniques significantly improve performance (percentage of evaluated segments) without decreasing reliability, and c) experience in assessing wall motion is the main factor in interobserver agr eement.