Quantitative proton magnetic resonance spectroscopy of the human cervical spinal cord at 3 tesla

Abstract

In vivo proton MR spectroscopy ( 1 H-MRS) is used extensively to study brain diseases and to add metabolic information to diagnostic MR examinations undertaken in routine clinical practice. The main central nervous system (CNS) metabolites detectable by 1 H-MRS are N-acetyl aspartate (NAA), creatine plus phosphocreatine (Cr), choline-containing compounds (Cho), and myoinositol (mI). Quantitative evaluation of these metabolites would be useful for the study of major spinal cord diseases, such as multiple sclerosis (MS), spinocerebellar ataxias, and intraaxial tumors, mainly because NAA may serve a neuronal marker function (1-4) in establishing axonal integrity (5-8), and mI signal change may be a marker of astrocytosis (9 -13).However, spinal cord proton spectroscopy is challenging because of technical difficulties that limit the quality of the spectroscopic data. The main drawbacks are the strong magnetic field inhomogeneities present in the spinal cord region, respiratory and cardiac movements, and the small size of the spinal cord. On a 3T system the increased signal-to-noise ratio (SNR) can help to find a good compromise between the size of the acquisition volume and the number of repetitions required to obtain a total acquisition time that is feasible in a clinical environment.Because of these technical challenges, few studies of spinal cord 1 H-MRS have been published to date. One study (14) quantified the main metabolites in six healthy volunteers using a 1.5T system (14). Another study involving MS patients and 12 healthy subjects was conducted (15), but the quantification data were not reported. A detailed study of the cervical spine proton spectroscopy protocol on a 2T system equipped with a purpose-built quadrature surface coil compared the concentrations of NAA, Cr, and mI on cervical spine and some brain areas in six healthy volunteers (16 -17). Two recent abstracts described studies on a 3T system: one involved a preliminary qualitative single-voxel comparison of the cervical and dorsal spine of a healthy volunteer and cervical spine tumor spectra (18), and one was a preliminary MR spectroscopic imaging (MRSI) quantification study (19).Our purpose was to devise an acquisition and postprocessing protocol to quantify the main CNS metabolites on the cervical spinal cord on a 3T system with standard equipment to include the spectroscopy study in routine clinical spinal cord imaging and define the normal metabolite concentrations in this location to use as a reference. This study presents a protocol for quantitative cervical spinal cord single-voxel MRS with the first mean relative concentrations for NAA, Cr, Cho, and mI in a group of 10 healthy volunteers using a clinical 3T system. These concentrations were also compared with brainstem metabolite contents in healthy subjects that were previously acquired with the same system. Statistical analysis confirmed a significant difference in metabolite concentrations between the spinal cord and the brainstem. MATERIALS AND METHODSSpinal cord 1 H-MRS examinatio...

Quantitative Cervical Spinal Cord 3T Proton MR Spectroscopy in Multiple Sclerosis

“…Automated optimization of gradient shimming, transmitter pulse power, and water suppression was used. To maintain acceptable acquisition times in a clinical setting and to compare data with the values obtained in healthy subjects, 27 we used automatic shimming and did not apply cardiac gating.…”

“…shtml) 27,31 (with LCMgui). As reported in the LCModel user manual, the program automatically recognizes the phased array data and combines them by a weighted average on the basis of the observed scatter in each channel.…”

“…The Student t test was used to evaluate the previously published 27 difference in healthy metabolite content of the cervical spinal cord, acquired with the same protocol on a group of 10 healthy volunteers (4 women and 6 men; range, 22-66 years of age; mean, 34.9 Ϯ 12.2 years) who were age-and sex-matched with the group of patients with RRMS, with a statistically significant threshold of P ϭ .01. The linear correlation between each metabolite ratio (tNAA:Cr, tNAA:Cho, Cho:Cr, and mIns:Cr) value and the EDSS scores for the patient group was evaluated by linear regression by using the statistical threshold P ϭ .01.…”

“…A recent study published a protocol for quantitative singlevoxel cervical spinal cord spectroscopy, presenting the first mean relative concentration ratios for NAA, creatine plus phosphocreatine (Cr), Cho, and mIns in a group of 10 healthy volunteers by using a clinical 3T MR imaging system. 27 The present study applied the same acquisition and postprocessing protocol to quantify the main CNS metabolites on the cervical spinal cord plaques of a group of 15 patients with relapsingremitting MS (RRMS) and compared them with the metabolite content of the group of healthy volunteers, to see if there was a difference between the MR spectroscopic findings of the cervical spinal cord in patients with RRMS and in volunteers. Then we evaluated whether these MR imaging measures correlate with disability in these patients, assessed through the Expanded Disability Status Scale (EDSS).…”

High field MRI in the diagnosis of multiple sclerosis: high field–high yield?

“…Moreover, 1 H-MRS at 3 T also offers chances to detect and quantify metabolic changes of CNS regions very important for diagnostic and prognostic purposes, which, however, are difficult to evaluate at lower field strengths because of lower SNR values and the need of high spatial resolution imaging such as the infratentorial brain and the spinal cord [80]. Diffusion tensor imaging Diffusion-weighted imaging in general and diffusion tensor imaging (DTI) in particular have been increasingly applied in the evaluation of pathological changes in MS.…”