Computer-assisted determination of left ventricular endocardial borders reduces variability in the echocardiographic

Eva Maret1, Lars Brudin, Lena Lindstrom

  • 1Department of Clinical Physiology, Ryhov County Hospital, SE-55185 Jonkoping, Sweden. eva.maret@bredband.net

Cardiovascular Ultrasound
|November 19, 2008
PubMed

Insights

This study shows that the corrected automatic echocardiographic method (AutoEF) improves left ventricular ejection fraction (LVEF) measurements, reducing variability for less experienced readers compared to manual methods.

Area of Science:

  • Cardiology
  • Medical Imaging
  • Echocardiography

Background:

  • Left ventricular size and function are critical prognostic indicators in heart disease.
  • Accurate measurement of left ventricular ejection fraction (LVEF) is essential for patient management but is operator-dependent.
  • Previous automated echocardiographic methods have faced limited adoption.

Purpose of the Study:

  • To evaluate an automated echocardiographic method (syngo AutoEF) for LVEF determination.
  • To compare the accuracy and efficiency of AutoEF against manual planimetry (Simpson's method) and visual assessment.
  • To assess the performance of AutoEF with both experienced and novice readers.

Main Methods:

  • Sixty patients undergoing myocardial perfusion imaging (MPI) were included.
  • Two-dimensional echocardiography was performed within one hour of MPI.
  • LVEF was analyzed using uncorrected AutoEF, corrected AutoEF, manual Simpson, and visual assessment, with comparisons to SPECT-derived LVEF.

Main Results:

  • Corrected AutoEF demonstrated reduced measurement variation compared to manual Simpson's method, particularly for novice readers.
  • Analysis time varied significantly across methods, with visual assessment being the fastest and corrected AutoEF taking longer than uncorrected AutoEF.
  • Agreement between methods was moderate when LVEF was categorized.

Conclusions:

  • The corrected AutoEF method enhances the reliability of LVEF measurements, decreasing inter-observer variability.
  • This automated approach is particularly beneficial for inexperienced echocardiography readers.
  • AutoEF offers a valuable tool for improving the consistency of LVEF assessment in clinical practice.
Abstract