Research pepars

Assessing Cardiac Dynamics based on X-Ray

Coronary Angiograms

Cardiac cycles is of importance for diagnosing coronary artery diseases and exploring pathogenesis of related circulation deficits and myocardial anomalies.

Specialists diagnose end systolic contour with end-diastolic contour of the left

Ventricle (LV) and measuring the ventricular volume. But various errors occurred in this performance, so we cannot be ensured this method.

When doing cardiac researchers it is rather difficult to quantitatively estimate the

deformation, such as expansion, contraction, and twisting, due to representations with spherical harmonic functions and they neglected variations of global location and orientation of the heart.so, bifurcation points extracted from in vivo angiograms is rather limited and they are sparsely distributed over the myocardia surface. Therefore, motion estimation results

from such limited data are not accurate enough.

Method

For this research used Parameters of each component are separately estimated based on non-rigid motion theory. And they followed following steps

1. Reconstructing 3-D Vessel Skeletons from Angiograms

The main advantage of this method is that matching between the

Angiogram pair in point-by-point manner is avoided.

2. Estimating Global Rigid Motion

In this quantitatively estimate global rigid motion.

3. Estimating Global Deformation

They first compensate the global rigid motion according to estimated parameters before analysing global deformation.

Human heart as a whole is believed to be an elastic body and its shape changes periodically over cardiac cycles. Therefore, we build a global shape model according to 3-D coronary vessel skeletons to estimate cardiac global deformations. In recent years, the

Deformable surface model has been widely used in the field of computer vision and image processing resulting in a globally smooth and coherent surface [8]. Especially,

Super quadrics (SQ) surface [9] has received more and more attentions because of its advantages of compact representation and robust recovery of 3-D objects.