NUMERICALLY OPTIMIZED EXPERIMENT DESIGN FOR MEASUREMENT OF GREY WHITEMATTER METABOLITE T-2 IN HIGH-RESOLUTION SPECTROSCOPIC IMAGES OF BRAIN/

T-2 relaxation measurements for choline (Cho), total creatine (Cr = cr eatine + phosphocreatine), and N-acetylaspartate (NAA) were made separ ately in eight healthy volunteers using an average of forty 0.5 cc vol umes (20 from grey matter and 20 from white matter) in spectroscopic i mages with a 32 x 32 resolution and a 240 mm field of view. In grey ma tter, the means and standard deviations of the T-2 values were 186 +/- 23, 149 +/- 10, and 232 +/- 15 ms for Cho, Cr, and NAA, respectively, and in white matter, the mean T-2 values were 178 +/- 16, 143 +/- 8, and 228 +/- 16 ms, respectively, with no significant differences betwe en grey and white matter. The high-resolution measurements of T-2 valu es were possible because of experimental planning based on the minimiz ation of predicted fitting uncertainties. Explicit expressions were de rived to estimate the uncertainties in T-2 values, and it was found th at two spectroscopic images with echo times of 50 and 250 ms, respecti vely, would yield sufficient precision for T-2 measurements. The deriv ation of the expressions, a discussion of their behavior, and the expe rimental planning and verification are presented. (C) 1995 Academic Pr ess, Inc.