On the reproducibility of hippocampal MEGA-sLASER GABA MRS at 7T using an optimized analysis pipeline

Yannik Völzke,Eberhard D Pracht,Elke Hattingen,Tony Stöcker,Desmond H Y.Tse

doi:10.1007/s10334-020-00879-9

Abstract

ObjectivesGABA is the most important inhibitory neurotransmitter. Thus, variation in its concentration is connected to a wide variety of diseases. However, the low concentration and the overlap of more prominent resonances hamper GABA quantification using MR spectroscopy. The hippocampus plays a pivotal role in neurodegeneration. Susceptibility discontinuities in the vicinity of the hippocampus cause strong B0 inhomogeneities, impeding GABA spectroscopy. The aim of this work is to improve the reproducibility of hippocampal GABA+ MRS.MethodsThe GABA+/total creatine ratio in the hippocampus was measured using a MEGA-sLASER sequence at 7 Tesla. 10 young healthy volunteers participated in the study. A dedicated pre-processing approach was established. Spectral quantification was performed with Tarquin. The quantification parameters were carefully adjusted to ensure optimal quantification.ResultsAn inter-subject coefficient of variation of the GABA+/total creatine of below 15% was achieved. Additional to spectral registration, which is essential to obtain reproducible GABA measures, eddy current compensation and additional difference artifact suppression improved the reproducibility. The mean FWHM was 23.1 Hz (0.078 ppm).ConclusionThe increased spectral dispersion of ultra-high-field spectroscopy allows for reproducible spectral quantification, despite a very broad line width. The achieved reproducibility enables the routine use of hippocampal GABA spectroscopy at 7 Tesla.

Highlights

GABA is the most important inhibitory neurotransmitter in the mammalian brain [1]
We investigate the reproducibility of GABA+ concentration measurements, resulting from MEGA-sLASER [13] experiments of the hippocampus
As a quality metric for the reproducibility, we investigate the inter-subject coefficient of variation of the GABA+/ total creatine ratio in a group of young healthy volunteers

Summary

Introduction

GABA is the most important inhibitory neurotransmitter in the mammalian brain [1]. GABA-MRS is hampered by its inherently low signal and the overlap of more prominent resonances. J-difference editing is often used to remove these overlapping resonances [5]. The GABA molecule contains three CH2 groups which resonate at 1.9, 2.3 and 3.0 ppm. Between these nuclear spins, J-coupling is present. J-difference editing makes use of this coupling to separate the GABA signal and the overlapping resonances. This is done by subtracting two spectra where the GABA signal undergoes different J-coupling evolution.

Objectives

Methods

Conclusion