Meeting Banner
Abstract #3924

Wavelet-Based Evaluation for the Thoracic Aorta Stiffness from CINE-MR Images

Abubakr El-Tahir1, Alain Lalande1, Marie Xavier1, Alexandre Cochet1, Nicolas Vignon1, Paul M. Walker1, Jean-Eric Wolf1, Franois Brunotte1

1LE2I, UMR CNRS 5158, Dijon, Burgundy, France

A new method is proposed to evaluate the aortic stiffness and to detect abnormalities in its condition. The methods relies on cine MR images to monitor the aortic vessel and to observe the evolution of its cross sectional area at a given site over time. The variation of the area is then analyzed using the wavelet decomposition. The wavelet spectrum is used to classify patients with aortic disorders from healthy subjects. This method was tested on patients with Marfan syndrome and MYH11 mutation. Promising results was obtained in comparison with the commonly used stiffness measurements like compliance and pulse-wave-velocity.

Keywords

abnormality absolute addition allows although analyzed aorta aortas aortic ascending attractive automatic axial burgundy cardiac carriers cine circulation class classes classical coefficients coll commonly compliance components condition confirms connective consideration considering consisted contour control correlation cross curve cycle decomposition descending determined developed diagnosing differentiate difficult direct disease diseases disorder disorders distinguish distinguishing domain done early effective encoded evaluate evaluation evolution fail finding findings frequency future gated gives hallmark healthy inadequate inherited insight invasive jean larger like localization magnitude material merely modality moreover mutation namely novel orientated overlap pathologies patient patients perpendicular plot population promising propose proposed pulmonary pulse recruited remains representation representing rest rose scatter segmentation separate separated serve sites spectral stiffness studied subject subjects suggests syndrome taken thoracic tissue track tracked transform transit trunk validated variation various velocity walker wave wavelet whole wide wolf