As the population of the developed world ages, and methods for cardiac therapy continually improve, there is a growing need for inexpensive, noninvasive assessment of cardiac function. Current strain rate imaging technology (SRI) has the potential to satisfy this need for a significant patient population since it can map contractile processes through and along the myocardial wall; however, its utility is limited because its estimates of deformation rate are angle-dependent. Pixel Velocity (PVI), working with the University of Michigan Biomedical Ultrasonics Lab (UM), proposes to develop an inexpensive correlation processor enabling high performance SRI of the heart. Our solution will provide real-time computation of tissue strain images and """"""""trashograms""""""""(i.e., images of the magnitude of the correlation coefficient.) The specific aims of this Phase I proposal are to specify the computational architecture and algorithm for: a) efficient, two-dimensional complex correlation of real-time ultrasound image sequences; b) SRI using adaptive frame decimation; and c) simultaneous real-time display of SRI and trashogram data. Finally, we will demonstrate the performance of candidate algorithms on a general purpose processor using real-time image data. The goal of Phase I is to provide hardware and software specifications to begin detailed design and implementation efforts in Phase II.
