Optical flow algorithm for cardiac motion estimation

Diagnostic techniques in cardiology require complex image analysis of single images and image sequences obtained by a variety of medical imaging modalities such as ECG-gated MR, CT, and ultrasound. Cardiac motion estimation provides useful information about cardiac function. Here, the authors present a novel brightness-based algorithm for computation of optical flow. The advantage of brightness-based techniques as opposed to shape-based techniques is that they do not require segmentation of images to obtain the shape of various heart regions. The proposed optical flow algorithm is applied to ECG-gated MR image sequence of the heart. The method used in this work is based on minimization of an energy function that is computed for two consecutive image frames. The energy function consists of two terms. The first energy term represents a measure of similarity between the regions in two image frames. The second energy term calculates the neighborhood influence and is increased based on the difference in velocities and the cross-correlation of neighboring regions. The regions are dynamically grown and are of variable size. The minimal energy state corresponds to the optimal solution of the optical flow estimation problem. The energy function minimization is based on a steepest descent algorithm. The proposed algorithm has been implemented in C language and tested on mathematical phantoms and real patient images. Experiments have shown encouraging results