Page Last Updated: May 17, 2026

Magnetic Resonance Spectroscopy (MRS)🔗

Magnetic Resonance Spectroscopy (MRS) measures biochemicals involved in neuronal metabolism, neurotransmission, and oxidative stress, enabling investigation of mechanisms underlying structural, functional, and behavioral development. HBCD is the first study of this scale to incorporate MRS into a comprehensive pediatric neuroimaging protocol, with data acquired via Integrated Short-TE and Hadamard Multi-Sequence (ISTHMUS) (Hui et al., 2024; Oeltzschner et al., 2019). Primary metabolites measured for HBCD include:

N-acetylaspartate (NAA): marker of neuronal mitochondrial metabolism
Glutamate and γ-aminobutyric acid (GABA): principal excitatory and inhibitory neurotransmitters
Glutathione (GSH): major antioxidant protecting against reactive oxygen species
Additional metabolites include lactate, ascorbate, creatine, myo-inositol, glutamine, and total choline

MRS Acquisition🔗

Full protocols, sequence installation, and operation instructions are available via HBCD Study MRI Protocols. Of note, a key challenge in MRS is sensitivity to scanner frequency drift, which can degrade data quality by altering editing efficiency and signal contributions (Harris et al., 2014). To address this, HBCD incorporates interleaved water referencing for real-time frequency correction, initially implemented on Philips systems and later extended to Siemens and GE platforms (Edden et al., 2016).

MRS Processing & Derivatives🔗

HBCD MRS data are processed with OSPREY-BIDS, a customized automated pipeline based on OSPREY (Oeltzschner et al., 2020; Zöllner et al., 2023). Full details regarding HBCD processing implementation (e.g., file selection for processing) are available in the external HBCD Processing documentation.

 How To Read File Trees →

hbcd/
└── derivatives/
    └── osprey/
        └── sub-[ID]/
            └── ses-[V0X]/
                ├── HERCULES/
                │   ├── PreOspreyLocalizerReg/
                │   │   ├── {aal|c1|c2|c3}reference_seg_aligned_to_localizer.nii.gz
                │   │   ├── reference_{img|seg}_aligned_to_localizer.nii.gz
                │   │   ├── readme.txt
                │   │   ├── registration_summary.json
                │   │   └── transform_mat.npy
                │   │
                │   ├── QuantifyResults/
                │   │   ├── {diff1|diff2|sum}_AlphaCorrWaterScaledGroupNormed_Voxel_1_Basis_1.tsv
                │   │   ├── {diff1|diff2|sum}_{AlphaCorr|CSF|raw|TissCorr}WaterScaled_Voxel_1_Basis_1.tsv
                │   │   └── {diff1|diff2|sum}_{amplMets|tCr}_Voxel_1_Basis_1.tsv
                │   │
                │   ├── SegMaps/
                │   │   ├── sub-[ID]_ses-[V0X]_acq-hercules_svs.nii_space-scanner_Voxel-1_label-{ROI}.nii.gz
                │   │   └── TissueFractions_Voxel_1.tsv
                │   │
                │   ├── VoxelMasks/
                │   │   └── sub-[ID]_ses-[V0X]_acq-hercules_svs_space-scanner_mask.nii.gz
                │   │
                │   ├── LogFile.txt
                │   ├── QM_processed_spectra.tsv
                │   ├── subject_names_and_excluded.tsv 
                │   ├── SummaryMRSinMRS.md
                │   └── wrapper_settings.mat
                │
                └── unedited/
                # Mirrors HERCULES/ structure - only unique filenames are displayed below
                    ├── QuantifyResults/
                    │   ├── A_AlphaCorrWaterScaledGroupNormed_Voxel_1_Basis_1.tsv
                    │   ├── A_{AlphaCorr|CSF|raw|TissCorr}WaterScaled_Voxel_1_Basis_1.tsv
                    │   └── A_{amplMets|tCr}_Voxel_1_Basis_1.tsv
                    ├── SegMaps/
                    │   └── sub-[ID]_ses-[V0X]_acq-shortTE_svs.nii_space-scanner_Voxel-1_label-{ROI}.nii.gz
                    └── VoxelMasks/
                        └── sub-[ID]_ses-[V0X]_acq-shortTE_svs_space-scanner_mask.nii.gz

# Label Values Legend
ROI: CSF, GM, Tha, WM

Osprey MRS Output User Guide🔗

The primary of outcome variables from MRS data processed through the Osprey pipeline are metabolite concentrations. The ISTHMUS acquisition generates four spectra, each modeled separately using linear combination modeling with inclusive basis sets (since most metabolites may contribute, at least minimally, to all spectra). Each spectrum provides optimal quantification for a different subset of metabolites as follows:

Spectrum	Best-Quantified Metabolites	Table Name (in tabulated pipeline derivatives)
Short-TE Unedited	tNAA, tCr, tCho, mI, Glx, Scyllo	`img_osprey_unedited_A_TissCorrWaterScaled_Voxel_1_Basis_1`
HERCULES Sum	NAA, Glu, Gln	`img_osprey_HERCULES_sum_TissCorrWaterScaled_Voxel_1_Basis_1`
HERCULES Diff1	GABA+	`img_osprey_HERCULES_diff1_TissCorrWaterScaled_Voxel_1_Basis_1`
HERCULES Diff2	GSH, Lac, NAAG, PE	`img_osprey_HERCULES_diff2_TissCorrWaterScaled_Voxel_1_Basis_1`

Quantification Approaches🔗

MRS values are quantified relative to internal reference signals. Optimal reference for HBCD data has not yet been established, so several quantification methods are provided (see Harris et al., 2015):

Method	Recommended Use
Tissue-Corrected Water-Scaled Concentrations	Recommended default (`TissCorrWaterScaled` TSV files)
Alpha-corrected variants	Use only for GABA+
Metabolite-to-creatine ratios	Values normalized to total creatine (tCr), relatively stable across lifespan. tCR is not present in CSF and should only be used for tissue-based analysis.
Unscaled model amplitudes	Do not use for analysis

References🔗

Bottomley, P. A. (1987). Spatial localization in NMR spectroscopy in vivo. Annals of the New York Academy of Sciences, 508(1), 333–348. https://doi.org/10.1111/j.1749-6632.1987.tb32915.x

Edden, R. A. E., Oeltzschner, G., Harris, A. D., Puts, N. A. J., Chan, K. L., Boer, V. O., Schär, M., & Barker, P. B. (2016). Prospective frequency correction for macromolecule-suppressed GABA editing at 3T. Journal of Magnetic Resonance Imaging, 44(6), 1474–1482.. https://doi.org/10.1002/jmri.25304

Harris, A. D., Glaubitz, B., Near, J., John Evans, C., Puts, N. A. J., Schmidt-Wilcke, T., Tegenthoff, M., Barker, P. B., & Edden, R. A. E. (2014). Impact of frequency drift on gamma-aminobutyric acid-edited MR spectroscopy. Magnetic Resonance in Medicine, 72(4), 941–948. https://doi.org/10.1002/mrm.25009

Harris, A. D., Puts, N. A. J., & Edden, R. A. E. (2015). Tissue correction for GABA-edited MRS: Considerations of voxel composition, tissue segmentation, and tissue relaxations.Journal of Magnetic Resonance Imaging, 42(5), 1431–1440. https://doi.org/10.1002/jmri.24903

Hui, S. C. N., Murali-Manohar, S., Zöllner, H. J., Hupfeld, K. E., Davies-Jenkins, C. W., Gudmundson, A. T., Song, Y., Yedavalli, V., Wisnowski, J. L., Gagoski, B., Oeltzschner, G., & Edden, R. A. E. (2024). Integrated Short-TE and Hadamard-edited Multi-Sequence (ISTHMUS) for advanced MRS. Journal of Neuroscience Methods, 409(110206), 110206. https://doi.org/10.1016/j.jneumeth.2024.110206

Oeltzschner, G., Saleh, M. G., Rimbault, D., Mikkelsen, M., Chan, K. L., Puts, N. A. J., & Edden, R. A. E. (2019). Advanced Hadamard-encoded editing of seven low-concentration brain metabolites: Principles of HERCULES. NeuroImage, 185, 181–190. https://doi.org/10.1016/j.neuroimage.2018.10.002

Oeltzschner, G., Zöllner, H. J., Hui, S. C. N., Mikkelsen, M., Saleh, M. G., Tapper, S., & Edden, R. A. E. (2020). Osprey: Open-source processing, reconstruction & estimation of magnetic resonance spectroscopy data. Journal of Neuroscience Methods, 343(108827), 108827. https://doi.org/10.1016/j.jneumeth.2020.108827

Träber, F., Block, W., Lamerichs, R., Gieseke, J., & Schild, H. H. (2004). 1H metabolite relaxation times at 3.0 tesla: Measurements of T1 and T2 values in normal brain and determination of regional differences in transverse relaxation. Journal of Magnetic Resonance Imaging, 19(5), 537–545. https://doi.org/10.1002/jmri.20053

Zöllner, H. J., Davies-Jenkins, C. W., Lee, E. G., Hendrickson, T. J., Clarke, W. T., Edden, R. A. E., Wisnowski, J. L., Gudmundson, A. T., & Oeltzschner, G. (2023). Continuous automated analysis workflow for MRS studies. Journal of Medical Systems, 47(1), 69. https://doi.org/10.1007/s10916-023-01969-6