Ultrasound Cardiac Strain Imaging of Left Ventricle Using Lagrangian Tracking
C Ma*, T Varghese, University of Wisconsin, Madison, WIWE-E-134-11 Wednesday 2:00PM - 3:50PM Room: 134
Purpose: Lagrangian description of myocardial tissue deformation is essential for accurate regional strain estimation of left ventricular wall over time. Failure to couple the estimated displacement and strain information with the correct myocardial tissue structure will lead to erroneous displacement and strain estimation.
Methods: In this study, a method to obtain Lagrangian based deformation tracking is presented. Myocardial tissue is divided into a fixed number of pixels whose deformation is tracked over the cardiac cycle. An algorithm that utilizes a polar-grid generated between the estimated endocardial and epicardial contours for the left ventricle in cardiac short axis images is proposed to ensure Lagrangian description. Displacement estimations from consecutive radiofrequency frames were then mapped onto the polar grid to obtain a distribution of the cumulated displacement that is mapped to the polar grid over time.
Results: This method was validated against a finite element analysis (FEA) based canine heart model coupled to an ultrasound simulation program. Segmental analysis of the accumulated displacement and strain over cardiac cycles show excellent agreement between the ideal result obtained directly from the FEA model and our Lagrangian approach for deformation mapping. Eulerian based estimation results on the other hand, show significant deviation from the ideal result. A comparison of the displacement and strain estimated on parasternal short axis view on a healthy volunteer is also presented.
Conclusion: Lagrangian deformation tracking using a polar grid demonstrates accurate results when validated against finite element cardiac model. In addition to the cardiac application, this approach can also be applied to transverse scans of arteries.