Automated border detection on contrast enhanced echocardiographic images

James N. Kirkpatrick, Roberto M. Lang, Savitri E. Fedson, Allen S. Anderson, James Bednarz, Kirk T. Spencer*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

9 Scopus citations

Abstract

Background: Accurate determination of left ventricular ejection fraction (LV EF) is of paramount importance in the evaluation of patients with cardiovascular disease. Quantitative techniques for the automated calculation of EF exist however, the robustness of these techniques is dependent on adequate endocardial border definition and therefore are difficult to use in patients with limited images. We sought to combine the endocardial border enhancing effects of contrast echocardiography with an automated border detection technique to provide quantitative and accurate determination of LV EF. Methods: Thirty-nine consecutive patients referred to nuclear cardiology for EF determination underwent radionuclide angiography followed by echocardiographic imaging using prototype software that allowed automated border detection during contrast infusion. Results: Adequate LV cavity opacification with contrast was possible in 38/39 patients. The mean radionuclide EF was 50±16% (range 19-73). There was no statistically significant difference between the mean nuclear EF and averaged echocardiographically determined EF (51±18%). The mean bias was 0.6 with limits of agreement that were +15 and -14. Conclusion: This study demonstrated that prototype software successfully tracked the contrast enhanced endocardial border allowing accurate calculation of LV EF.

Original languageEnglish (US)
Pages (from-to)164-167
Number of pages4
JournalInternational Journal of Cardiology
Volume103
Issue number2
DOIs
StatePublished - Aug 18 2005

Keywords

  • Border detection
  • Contrast agents
  • Ejection fraction
  • Transthoracic echocardiography

ASJC Scopus subject areas

  • Cardiology and Cardiovascular Medicine

Fingerprint Dive into the research topics of 'Automated border detection on contrast enhanced echocardiographic images'. Together they form a unique fingerprint.

Cite this