Accuracy & Reproducibility of Short-TE MRS Measurements of GABA at 3T as a Function of Linewidth & SNR

Jamie Near¹, Jesper Andersson¹, Philip Cowen², Peter Jezzard¹

We investigate the accuracy and reproducibility of short-TE MRS measurements of GABA as a function of changing experimental conditions (linewidth and SNR). This is achieved by generating a large number of simulated datasets with known metabolite concentrations. Each dataset was then analysed using LCModel, and the estimated GABA concentrations were compared with the actual known values. This procedure was repeated for a range of experimental conditions, and the reproducibility and accuracy were calculated for each. Under experimental conditions corresponding to ACC and PCC, the reproducibility errors were 17% and 10%, respectively, and the estimation errors were 16% and 3%, respectively.