Imagent improves endocardial border delineation, inter-reader agreement, and the accuracy of segmental wall motion assessment

Navin C. Nanda, Dalane W. Kitzman, Howard C. Dittrich, Gail Hall, Massood Ahmad, Gerard Aurigemma, Leon Axel, Benjamin Byrd, Farooq Chaudhry, Jerald Cohen, Harold Copans, Linda Crouse, George W. Dennish, Kenneth Ford, Linda Gillam, Martin Goldman, Paul Grayburn, David Hill, Michael Imburgia, Larry JacobsMorris Kotler, George Lighty, Sunil Mankad, Natesa Pandian, John Panidis, Julio Perez, Joel Raichlen, David Rubenson, Thomas Ryan, Douglas Segar, Edward Shapiro, Jamshid Shirani, Martin St John Sutton, Stephen A. Stowers, William Van Decker, Flordeliza Santos Villanueva, William Zoghbi

Research output: Contribution to journalArticlepeer-review

25 Scopus citations


Background: This study was conducted to assess the ability of a new echocardiographic contrast agent, Imagent (perflexane lipid microspheres; Alliance Pharmaceutical Corp., San Diego, CA), to improve endocardial border delineation (EBD) and assessment of segmental wall motion (SWM). This was achieved by analysis of inter-reader agreement by echocardiography and comparison with an independent imaging technique, magnetic resonance imaging (MRI). Methods: Two separate, independent, prospective, randomized, controlled, multicenter trials were conducted at 26 centers and included a total of 409 efficacy-evaluable patients. In Study A 206 patients were randomized to receive either Imagent or saline and in Study B, 203 patients received Imagent with a subset imaged by both echocardiography and MRI. All patients were required to have suboptimal baseline images using fundamental imaging. Images were optimized at baseline prior to contrast and the settings maintained post-contrast. Imagent, a suspension of perfluorohexane-filled spheres with flexible lipid shells, was administered as an IV bolus at 0.125 mg/kg body weight. Gated MRI studies were performed within 48 hours of dosing in a subset of 26 subjects. Six expert independent blinded readers reviewed unpaired noncontrast and contrast exams and scored EBD and SWM. Analysis of interreader agreement was performed by comparing the SWM score (1 to 5) recorded by each reader pair. In addition, unanimity between readers for SWM was evaluated. For comparison to MRI, the results from echo readers 4, 5, and 6 were each compared with a single independent MRI reader. Results: The patients enrolled in these clinical trials displayed markedly suboptimal images with 49% and 71% (Study A and Study B) of the segments determined by the readers to be suboptimal prior to contrast administration. All readers recorded statistically significant (P < 0.0001) improvement in total EBD scores following the administration of Imagent. Comparison of noncontrast SWM scores for each pair of echo readers resulted in agreement in an average of 39%, of segments in Study A, and 31% of the segments in Study B. Use of Imagent improved agreement in SWM scores to 65% in Study A, and 48% of segments in Study B (P < 0.0001) for all reader pairs in both studies. Reader unanimity in SWM scores increased from 13% to 41% of the segments with the administration of Imagent. Blinded review of the noncontrast echo examinations resulted in agreement with MRI derived SWM scores in 15% of the segments. The administration of Imagent improved this agreement to 47%, of the segments (P ≤ 0.0005 for each blinded reader). Conclusions: Use of Imagent during echocardiographic imaging improves EBD, providing a significant improvement in inter-reader agreement in SWM evaluation with echo and greater than a threefold improvement in SWM scoring accuracy with MRI.

Original languageEnglish (US)
Pages (from-to)151-161
Number of pages11
Issue number2
StatePublished - Feb 2003


  • Contrast echocardiography
  • Imagent
  • LV endocardial border delineation

ASJC Scopus subject areas

  • Cardiology and Cardiovascular Medicine
  • Radiology Nuclear Medicine and imaging


Dive into the research topics of 'Imagent improves endocardial border delineation, inter-reader agreement, and the accuracy of segmental wall motion assessment'. Together they form a unique fingerprint.

Cite this