Global Optimization of Optical Flow Technique on the Computation of Tissue-motion in Neonatal Cranial Ultrasonogram

In this paper, a global optimization method of optical flow technique is presented for estimating tissue-motion velocity from a time series of neonatal cranial ultrasonogram images. The global optimization method is used here was proposed by Horn and Schunck. In this method, the tissuemotion velocity field is determined by minimizing an error function based on the gradient constraint and a global smoothing term over the whole image. An iterative implementation is shown which successfully computes the tissue-motion velocity for a number of cranial ultrasonogram image sequences. Two dimensional tissue-motion vectors are presented, shows that artery pulsation can be detected in low brightness image. It is shown that, the strength of artery pulsation can be quantitatively estimated by using the magnitude of tissue motion vector.