3D Optical Flow Methods in Cardiac Imaging

Heart disease is a leading cause of death in the Western world and, as a result, the study of heart behaviour is becoming increasingly important. In the UK coronary heart disease is the most costly disease and is the most common cause of death. Transposing research results of digital image analysis into the medical domain is not a new idea. This paper discusses techniques for motion analysis in cardiac images. The quantiication of motion results from such images can give insight into the hearts health and events aaecting its performance. The aim is to develop a system, to compute 3D volumetric (optical) ow and, from this, assist in the identiication of heart wall motion abnormalities which will lead to diagnosis of heart diseases. This would allow the recovery of cardiac motion with parameter extraction whose signiication would be easier for cardiologists to understand than previous 3D analysis. Firstly, a review is carried out in the area of cardiac motion analysis. The computation of 3D volumetric optical ow on gated MRI datasets is discussed. The well known 2D least squares and regularization approaches of Lucas and Kanade 13 and Horn and Schunck 11 are extended. Flow elds are shown (as XY and XZ 2D ows) for a beating heart. The ow results not only can capture the expansion and contraction of various parts of the heart motion but also can capture the twisting motion of the heart.