The measurement of flow and motion in vivo is a powerful clinical tool. Currently, angiography, including digital subtraction angiography, and Doppler ultrasound are used to examine the blood vessels; cineangiography, multigated radionuclide studies, gated X-ray CT, and Doppler ultrasound are used to assess the motion of the heart. Magnetic resonance (MR) imaging is an attractive alternative to these techniques because it is non-invasive, has good spatial resolution, and, most importantly, it is inherently sensitive to the motion of resonant nuclei. Techniques of proton nuclear magnetic resonance imaging (MRI) will be developed to image motion. Motion images are two- or three-dimensional spatial maps of one or more parameters of motion. For exampole, a two-dimensional velocity image is one in which the intensity of each pixel reflects the speed and direction of motion in the volume defined by that pixel and the image slice. Specific parameters to be determined are the type of motion (e.g., laminar or turbulent) and velocity, acceleration, and jerk (the time rate of change of acceleration) both in the plane of the image and transverse to it. Emphasis will be placed on developing fast MR motion imaging techniques because the data collection time for an MR image is already long compared to X-ray CT. Software simulations will be written to foster an understanding of the problem and to provide data for the development and testing of the analytical software for phantom and in vivo data. Flow and motion phantoms will be built and imaged to test the imaging techniques. These phantoms will allow the isolation of the parameters of flow or motion to be tested. After verification of the techniques using phantoms, dogs will be imaged. Blood vessels and cardiac wall motion will be assessed by MRI and verified by ultrasound techniques to assess the performance of the MR flow and motion imaging techniques in vivo. The long-term goal of this research is to develop fully the potential of MRI to parameterize blood vessel flow and cardiac wall motion with greater resolution, in more detail, and with less discomfort to the patient than is provided by currently employed non-MR techniques. Our preliminary work with software simulations and with imaging phantoms confirms that this is a realistic goal.
