In FY '97, LDRR was transferred from the Office of Director, NIH to the Clinical Center. Significant improvements and developments in fast scanning techniques for functional magnetic resonance (MR) imaging of the brain were achieved by the addition of new prototypic scaleable gradient amplifiers to the MR unit. For Echo planar imaging and spiral scanning, the prototype system increased the duty cycle from 25 to 50 percent, resulting in a twofold improvement in scan speed. These improvements allowed our group to develop and test a new echo-shifted trapezoidal spiral sequence that acquired 30 slice locations/second (compared with 18 slices/second for single-shot spiral imaging) with comparable T2* weighting and temporal shot-to-shot stability to other single-shot techniques. This will allow LDRR to perform signal event based functional MRI studies or dynamic contrast studies of the whole head at 1-second temporal resolution. This will provide valuable information for the timing of activation and neural connections between activated regions of the central nervous system. In addition, post-processing strategies were evaluated and published that allowed for effective and straightforward correction of amplifier nonlinearities. These strategies are currently being implemented as a standard feature for all our fast scanning studies.This significant new development in proton spectroscopic imaging (MRSI) involves the acquisition of metabolic data without suppression of the water signal using a multislice long echo time sequence. MRSI without water suppression uses the high dynamic range that recently became available with state-of-the-art MRI instrumentation. This new technique will allow for reliable quantitation of the concentration of brain metabolites by using the water signal as a reference. With this approach, corrections can be made to effect the instrumental variation and subject motion on the measured metabolite intensities. In addition to the hardware demands, it requires dedicated post-processing techniques of the data, involving reliable time-domain fitting routines. A prototype software spectroscopic imaging analysis package has been developed, which has demonstrated robust operation on normal volunteers. These developments should lead to the detection of Choline, Creatine, N-Acetyl aspartate and Lactate, with improved sensitivity and improved reliability. This is expected to be important in diagnosis and treatment planning, as well as the characterization of a number of diseases and neurologic abnormalities.
