Abstract
BACKGROUND
The optimization of magnetic resonance spectroscopy (MRS) sequences allows improved diagnosis and prognosis of neurological and psychological disorders. Thus, to assess the test-retest and intersequence reliability of such MRS sequences in quantifying metabolite concentrations is of clinical relevance.
PURPOSE
To evaluate the test-retest and intersequence reliability of three MRS sequences to estimate GABA and Glx = Glutamine+Glutamate concentrations in the human brain.
STUDY TYPE
Prospective.
SUBJECTS
Eighteen healthy participants were scanned twice (range: 1 day to 1 week between the two sessions) with identical protocols.
FIELD STRENGTH/SEQUENCE
3T using a 32-channel SENSE head coil in the PCC region; PRESS, JPRESS, and MEGA-PRESS sequences.
ASSESSMENT
Metabolite concentrations were estimated using LCModel (for PRESS and MEGA-PRESS) and ProFit2 (for JPRESS).
STATISTICAL TESTS
The test-retest reliability was evaluated by Wilcoxon signed-rank tests, Pearson's r correlation coefficients, intraclass-correlation coefficients (ICC), coefficients of variation (CV), and by Bland-Altman (BA) plots. The intersequence reliability was assessed with Wilcoxon signed-rank tests, Pearson's r correlation coefficients, and BA plots.
RESULTS
For GABA, only the MEGA-PRESS sequence showed a moderate test-retest correlation (r = 0.54, ICC = 0.5, CV = 8.8%) and the BA plots indicated good agreement (P > 0.05) for all sequences. JPRESS provided less precise results and PRESS was insensitive to GABA. For Glx, the r and ICC values for PRESS (r = 0.87, ICC = 0.9, CV = 2.9%) and MEGA-PRESS (r = 0.70, ICC = 0.7, CV = 5.3%) reflect higher correlations, compared with JPRESS (r = 0.39, ICC = 0.4, CV = 20.1%).
Baeshen, Arwa