Metabolite Quantification

Localized proton magnetic resonance spectroscopy (H MRS) became an important tool for investigating brain metabolism non-invasively. In animal models, high field H MRS allows detection of almost 20 metabolites [1-4], also called the neurochemical profile (taurine (Tau), creatine (Cr), phosphocreatine (PCr), phosphocholine (PCho), glycerophosphocholine (GPC), glutamate (Glu), glutamine (Gln), myo-inositol (Ins), ascorbate (Asc), alanine (Ala), aspartate (Asp), γ-aminobutyrate (GABA), glutathione (GSH), N-acetylaspartate (NAA), Nacetylaspartylglutamate (NAAG), glucose (Glc), lactate (Lac), phosphoethanolamine (PE) and glycine (Gly). Since brain metabolite concentrations can vary depending on the type of pathology, the usefulness of H MRS has been demonstrated in many brain disorders, e.g., in hepatic encephalopathy, Alzheimer’s, Huntington’s, and Parkinson’s diseases, acute traumatic brain injury, cancer, dementia, etc. [5-6]. The metabolite changes help to understand metabolic processes related to the studied pathologies and/or enable to monitor effect of treatment. At longer echo times (TE above 20 ms), the number of spectral lines, which can be used for quantification, is reduced. As a result, in most in vivo H MRS studies only changes of NAA, Cr+PCr, Cho and Lac have been detected. Short-echo-time in vivo H MRS spectra (TE=1-20 ms) contain more information due to minimal distortions of multiplets of coupled spin systems such as Glu, Gln, Ins, Glc, Asp, Ala, GABA, Asc, PE and Tau. Accurate and precise quantification of brain metabolites is challenging and depends on several parameters: hardware performance, pulse sequence design, adjustments of acquisition parameters, data processing and quantification strategies. The purpose of data processing is to estimate the signal amplitude or peak area of each metabolite in a given spectrum, which is proportional to the metabolite concentration. Then by using different quantification methods the signal amplitude or peak area is converted in tissue metabolite tissue concentration. Measurements at high magnetic field benefit from increased spectral dispersion and higher signal to noise ratio, which likely improve quantification precision and accuracy [1, 7-8]. However, due to overlap of proton signals from brain metabolites, sophisticated approaches for the separation of the metabolite signals are required. Consequently, metabolite concentrations are usually determined by fitting a measured in vivo H MRS spectrum to a linear combination of spectra of individual metabolites (also called metabolite basis set). The metabolite basis set can be obtained either by measuring aqueous solutions of pure metabolites or by quantum-mechanical simulations using known spectral parameters [9]. In the present course an overview of the factors that should be considered when performing and evaluating metabolite quantification is given. The course will be structured in 6 sections: 1. Short description of the time domain and frequency domain algorithms. 2. Factors affecting the quality and reliability of in vivo spectra 3. Key points to consider during the quantification (lineshape correction, macromolecule contribution) 4. Quantification techniques