تتبع الحركة القلبية في صورة الرنين المغنطيسي باستخدام تسجيل الصور اللاخطي

الملخص EN

The process of tracking the heart motion helps the physicians to locate the actual incidence of myocardial disease like ischemia and on the possible risk degree of the affected area of the heart.

In the present work, three different techniques of nonlinear image registration were implemented and applied to cardiac Magnetic Resonance Images (MRI) to track the heart motion, which is fast and small amplitude.

The first of these techniques depends on Demons algorithm that is based on optical flow principle and the second approach based on applying control points by using the finite elements method to track the location change displacements of these points, the third technique depends on B-Splines curves and the displacements are calculated by Free Form Deformations (FFD) approach.

The sum of squared difference (SSD) has been adopted as a similarity metric for the registration methods mentioned above to calculate the displacements of heart motion that is being used in cases of registering the mono-modality images and in case of Gaussian noise as in MR images.

It has shown that the registration method based on B-Spline has given the best results and can be used efficiently for MR images registration and thus calculate the heart displacement.