Premium
Calculation of three‐dimensional left ventricular strains from biplanar tagged MR images
Author(s) -
Moore Christopher C.,
O'Dell Walter G.,
McVeigh Elliot R.,
Zerhouni Elias A.
Publication year - 1992
Publication title -
journal of magnetic resonance imaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.563
H-Index - 160
eISSN - 1522-2586
pISSN - 1053-1807
DOI - 10.1002/jmri.1880020209
Subject(s) - interpolation (computer graphics) , tracking (education) , cardiac cycle , physics , point (geometry) , cardiac magnetic resonance , magnetic resonance imaging , human heart , motion (physics) , artificial intelligence , computer science , computer vision , algorithm , mathematics , geometry , medicine , radiology , cardiology , psychology , pedagogy
The noninvasive measurement of time‐resolved three‐dimensional (3D) strains throughout the myocardium could greatly improve the clinical evaluation of cardiac disease and the ability to mathematically model the heart. On the basis of orthogonal arrays of tagged magnetic resonance (MR) images taken at several times during systole, such strains can be determined, but only after heart motion through the image planes is taken into account. An iterative material pointtracking algorithm is presented to solve this problem. It is tested by means of mathematical models of the heart with cylindric and spherical geometries that undergo deformations and bulk motions. Errors introduced by point‐tracking interpolation were found to be negligible compared with those due to marker identification on the images. In a human heart studied with this technique, the corrected radial strains at the left ventricular base were approximately 2.5 times the two‐dimensional estimates derived from the fixed image planes. The authors conclude that material point tracking allows accurate, timeresolved 3D strains to be calculated from tagged MR images, and that prior correction for motion of the heart through image planes is necessary.