SHORT TE H-1 SPECTROSCOPIC MR-IMAGING OF NORMAL HUMAN BRAIN - REPRODUCIBILITY STUDIES

Studies have been performed to evaluate the reproducibility of longitu dinal acquisitions with short TE spectroscopic magnetic resonance (MR) imaging of human brain. In healthy volunteers, the ratios of N-acetyl -aspartate (NAA) and choline (Cho) with respect to total creatine (Cr) (creatine and phosphocreatine) have been examined in terms of voxel-t o-voxel, acquisition-to-acquisition, and subject-to-subject variation. Overall coefficients of variation for NAA/Cr ratios and Cho/Cr ratios were 18% and 16%, respectively. The interacquisition variation was no t statistically significant (P = .05) for either ratio. In terms of F ratios, the dominant source of variation in the NAA/Cr ratio was inter subject variability, whereas for the Cho/Cr ratio the dominant source of variation was intervoxel variability. Consequences of spectroscopic processing with spatial apodization and postacquisition water suppres sion and resolution enhancement were also examined. Such reproducibili ty studies are crucial if longitudinal spectroscopic evaluations are t o be performed in a clinical setting to follow the natural course of d isease states or the effects of therapy.