Geometric regularization improves 2D myocardial motion estimates in the mouse : an in-silico study

Estimation of myocardial deformation in the mouse based on speckle tracking using real time ultrasound imaging is feasible but remains challenging. Indeed, due to the low frame rate (about 300 Hz) in comparison to the heart rate (500-700 bpm), subsequent images show important decorrelation noise. As such, regularization of the motion estimates is essential. The objective of this work was to analyze the effect of different geometric regularization methods on the quality of the estimates in a theoretical setting. Estimated motion was regularized and accumulated from end-diastole until end-systole. From the results, it appears that limiting the spatial curvature of the resulting vector field is the best geometric regularizer. © 2009 Springer Berlin Heidelberg.Kremer F., Choi H.F., Langeland S., D’Agostino E., Claus P., D’hooge J., ''Geometric regularization improves 2D myocardial motion estimates in the mouse : an in-silico study'', IFMBE proceedings 4th European conference of the International Federation for Medical and Biological Engineering, vol. 22, pp. 555-558, November 23-27, 2008, Antwerp, Belgium.status: publishe