In the first three year period of this grant we have studied many of the aspects of Magnetic Resonance (MR) image synthesis. This has included optimization of the acquisition protocol, validation of synthetic images, development of high speed instrumentation for performing synthesis, and initiation of clinical studies. We next propose first, to continue the technical and clinical development of image synthesis, and second, to apply the instrumentation to high speed MR image acquisition. The individual parts of this proposal are: 1. MR Image Synthesis. Technical refinements will be made in image acquisition and manipulation. A pulse sequence will be developed enabling dual-TR multi-slice multi-echo acquisition with respiratory compensation. This will expedite the acquisition of MR images of the body. Weighted linear image combination will be studied as a means for maximizing contrast-to-noise (CNR) from a set of acquired images. Modifications of this method will enable selective material cancellation. Image synthesis will be evaluated in specific clinical applications of meningioma, liver, carcinoma of the bladder, and tumor and infection of the bone marrow. Synthesis will be used to predict optimized pulse sequences. 2. Imaging Instrumentation. Techniques developed for high speed image synthesis will be applied to high speed realtime image reconstruction. This will be applicable to recent acquisition schemes which enable scan times of 2-10 seconds. The technique of "MR Fluoroscopy" will be developed in which modestly degraded images will be displayed in realtime during acquisition at 2 image/second rates. In later years the instrumentation will be further developed and applied to quasi-realtime MR angiography and three-dimensional Fourier transform acquisition techniques.