2021•NMR in BiomedicineRequires access

Quantification of magnetic resonance spectroscopy data using a combined reference: Application in typically developing infants

Ryan J. Larsen, Borjan A. Gagoski, Sarah U. Morton, Yangming Ou, Rutvi Vyas, Jonathan S. Litt, P. Ellen Grant, Bradley P. Sutton

Open publisher page 10 citations

Abstract

Quantification of proton magnetic resonance spectroscopy (1H‐MRS) data is commonly performed by referencing the ratio of the signal from one metabolite, or metabolite group, to that of another, or to the water signal. Both approaches have drawbacks: ratios of two metabolites can be difficult to interpret because study effects may be driven by either metabolite, and water‐referenced data must be corrected for partial volume and relaxation effects in the water signal. Here, we introduce combined reference (CRef) analysis, which compensates for both limitations. In this approach, metabolites are referenced to the combined signal of several reference metabolites or metabolite groups. The approach does not require the corrections necessary for water scaling and produces results that are less sensitive to the variation of any single reference signal, thereby aiding the interpretation of results. We demonstrate CRef analysis using 202 1H‐MRS acquisitions from the brains of 140 infants, scanned at approximately 1 and 3 months of age. We show that the combined signal of seven reference metabolites or metabolite groups is highly correlated with the water signal, corrected for partial volume and relaxation effects associated with cerebral spinal fluid. We also show that the combined reference signal is equally or more uniform across subjects than the reference signals from single metabolites or metabolite groups. We use CRef analysis to quantify metabolite concentration changes during the first several months of life in typically developing infants.

About this research paper

What this paper is about

Quantification of proton magnetic resonance spectroscopy (1H‐MRS) data is commonly performed by referencing the ratio of the signal from one metabolite, or metabolite group, to that of another, or to the water signal. Both approaches have drawbacks: ratios of two metabolites can be difficult to interpret because study effects may be driven by either metabolite, and water‐referenced data must be corrected for partial volume and relaxation effects in the water signal. Here, we introduce combined reference (CRef) analysis, which compensates for both limitations. In this approach, metabolites are referenced to the combined signal of several reference metabolites or metabolite groups. The approach does not require the corrections necessary for water scaling and produces results that are less sensitive to the variation of any single reference signal, thereby aiding the interpretation of results. We demonstrate CRef analysis using 202 1H‐MRS acquisitions from the brains of 140 infants, scanned at approximately 1 and 3 months of age. We show that the combined signal of seven reference metabolites or metabolite groups is highly correlated with the water signal, corrected for partial volume and relaxation effects associated with cerebral spinal fluid. We also show that the combined reference signal is equally or more uniform across subjects than the reference signals from single metabolites or metabolite groups. We use CRef analysis to quantify metabolite concentration changes during the first several months of life in typically developing infants.

Why it matters

OpenAlex reports 10 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Quantification of proton magnetic resonance spectroscopy (1H‐MRS) data is commonly performed by referencing the ratio of the signal from one metabolite, or metabolite group, to that of another, or to the water signal. Both approaches have drawbacks: ratios of two metabolites can be difficult to interpret because study effects may be driven by either metabolite, and water‐referenced data must be corrected for partial volume and relaxation effects in the water signal. Here, we introduce combined reference (CRef) analysis, which compensates for both limitations. In this approach, metabolites are referenced to the combined signal of several reference metabolites or metabolite groups. The approach does not require the corrections necessary for water scaling and produces results that are less sensitive to the variation of any single reference signal, thereby aiding the interpretation of results. We demonstrate CRef analysis using 202 1H‐MRS acquisitions from the brains of 140 infants, scanned at approximately 1 and 3 months of age. We show that the combined signal of seven reference metabolites or metabolite groups is highly correlated with the water signal, corrected for partial volume and relaxation effects associated with cerebral spinal fluid. We also show that the combined reference signal is equally or more uniform across subjects than the reference signals from single metabolites or metabolite groups. We use CRef analysis to quantify metabolite concentration changes during the first several months of life in typically developing infants.

Key concepts: Metabolite, SIGNAL (programming language), Nuclear magnetic resonance, T2 relaxation, Chemistry, Magnetic resonance imaging, Biological system, Physics

Related papers

Back to paper searchBrowse research topicsOriginal source
Quantification of magnetic resonance spectroscopy data using a combined reference: Application in typically developing infants — Research Paper | ScholarLens