A noninvasive comparative study of myocardial strains in ischemic canine hearts using tagged MRI in 3-D

H. Azhari, J. L. Weiss, W. J. Rogers, C. O. Siu, E. P. Shapiro

Research output: Contribution to journalArticlepeer-review


Regional reduction in myocardial function has long been utilized for noninvasive detection and localization of ischemic regions in the left ventricle. In this study, we compared the ability of 13 different myocardial strain indexes to discriminate between ischemic and nonischemic tissue. Using magnetic resonance imaging with tagging and three-dimensional reconstruction, we noninvasively tracked the deformation of 24 myocardial cuboids in each of nine canine hearts with acute ischemia induced by coronary artery ligation. Monastral blue staining was used for postmortem assessment of ischemia. The ability of each index to discriminate between normal and ischemic zones was evaluated using the 't' statistic. The best discrimination was obtained by the endocardial area strain [t = 7.5 ± 3.0 (SD)], a local measure of systolic shrinkage on the endocardial surface, and by the magnitude of the endocardial principal strain (t = 7.0 ± 1.9). Transmural wall thickening (t = 6.1 ± 2.3) was the third best functional index. The same three indexes correlated most closely with perfusion, as assessed by monastral blue staining. These findings imply that these are the preferable functional indexes for mapping of ischemic myocardial regions.

Original languageEnglish (US)
Pages (from-to)H1918-H1926
JournalAmerican Journal of Physiology - Heart and Circulatory Physiology
Issue number5 37-5
StatePublished - 1995


  • ischemia
  • left ventricle
  • magnetic resonance imaging
  • magnetic resonance imaging tagging
  • three dimensional

ASJC Scopus subject areas

  • Physiology
  • Cardiology and Cardiovascular Medicine
  • Physiology (medical)


Dive into the research topics of 'A noninvasive comparative study of myocardial strains in ischemic canine hearts using tagged MRI in 3-D'. Together they form a unique fingerprint.

Cite this