Aims: The aim of this study was to assess the incremental value of tissue harmonic imaging vs conventional echocardiography for evaluating left ventricular ejection fraction by manual and automated quantitation as well as visual estimation in patients with distorted left ventricles.
Methods and results: In 25 patients unselected for image quality and with distorted left ventricles who underwent a nuclear study, digital cineloops of standard apical views were acquired by both tissue harmonic imaging and conventional echocardiography and sent to six observers for analysis of visual and quantitative left ventricular ejection fraction. Tissue harmonic imaging improved both the correlation and agreement of all echo techniques with nuclear measures, compared with conventional echocardiography echo, reducing standard errors (SE) to below 10%: for the visual estimate SE=7.5%, for manual tracing SE=6.3% and for automated tracing SE=8%. Tissue harmonic imaging decreased inter-observer variability compared with conventional echocardiography echo for both visual assessment (12.4% vs 18.4%, P<0.05) and quantitative measures (for manual tracing, 8.2% vs 11.8%, P<0.05; for automated tracing, 7.8% vs 16.8%, P<0.05).
Conclusions: In patients with distorted left ventricles unselected for image quality, tissue harmonic imaging improves accuracy and reproducibility of both visual and quantitative echocardiographic assessment of left ventricular ejection fraction. In particular, it promotes automated quantitation by reducing its high standard error into a clinically reasonable range.