Spectroscopy of the brain

Itamar Ronen, Hermien E. Kan, Chloé Najac, Andrew G. Webb


Abnormal cellular metabolism is commonly observed in many brain disorders, including neurodegenerative diseases (Alzheimer’s disease), autoimmune diseases (Multiple Sclerosis or Systemic Lupus) and cancer. These changes usually begin very early on and precede macroscopic structural damage. Thus, they are often invisible by magnetic resonance imaging (MRI). While conventional MRI is primarily used to generate anatomical images, magnetic resonance spectroscopy (MRS) allows to quantify cellular metabolism in the brain and other parts of the body, including heart, liver, muscles (see webpage on MRS and MRI in muscle diseases). It is therefore commonly used to identify metabolic changes associated with brain disorders and monitor treatment response. Data can either be obtained from one region of interest (so-called single voxel-of-interest) or multiple regions (whole brain) in the case of spectroscopic imaging.

Proton (1H) MRS is the most common method to quantify neurochemical changes. The number of quantifiable endogenous metabolites depends on the RF pulse sequence, the spectral resolution and signal-to-noise ratio. At ultrahigh field (7 Tesla), the signal-to-noise and spectral resolution are increased resulting in higher sensitivity and specificity. Therefore, up to 17 endogenous metabolites can be detected. Neurons and glial cells are the main cells of the central nervous system. The cell-specific compartmentation of some metabolites allows to specifically study neuronal and glial alterations by 1H MRS. For instance, N-acetyl-aspartate and glutamate are primarily located in neurons, whereas choline and myo-inositol are mainly found in astrocytes (see illustration below). In multiple sclerosis, a decrease in NAA levels has been associated with axonal damage whereas an increase in myo-inositol has been correlated with gliosis (i.e. proliferation and hypertrophy of glial cells). In cancer, an increase in choline levels is a biomarker of cell proliferation and tumor aggressiveness. The MRS data can also be modulated to gain access to tissue microstructure (see webpage on Diffusion Weighted Spectroscopy) or brain function and metabolism. As opposed to functional MRI (fMRI) which gives an indirect measure of neuronal activity based on the blood oxygenation level difference (BOLD effect), functional proton (1H) MRS (fMRS) provides a direct measurement of metabolic and cellular changes tightly associated with neuronal activation. Task-related changes (e.g. visual stimulation) in metabolites’ level are measured dynamically and can be correlated to activation (neurotransmission, synaptic plasticity) in specific brain regions. Beyond proton spectroscopy, other X nuclei can also be studied using MRS. In particular, 31P MRS gives access to important metabolic information. For instance, the rate of adenosine triphosphate (ATP) synthesis, the molecule that fuels important cellular metabolic reactions in our body, and intracellular pH can be measured.


At the C.J. Gorter Center, we are particularly interested in taking advantage of the ultrahigh field Philips MR scanner (7 Tesla) to:
  1. Develop original and robust methods for single voxel-of-interest and whole brain MRS acquisitions
  2. Study brain activity using fMRS
  3. Assess metabolic changes associated with brain disorders using 1H MRS, 31P MRS, fMRS  


  • University of Minnesota, USA
  • Danish Research Center for Magnetic Resonance, Denmark
  • Weizmann Institute, Israel
  • Amsterdam Medical Center, The Netherlands
  • Utrecht Medical Center, The Netherlands


Selected publications

  • N. Doorenweerd et al., Proton Magnetic Resonance Spectroscopy Indicates Preserved Cerebral Biochemical Composition in Duchenne Muscular Dystrophy Patients, Journal of Neuromuscular Diseases, 2017;4(1):53-58
  • Zielman R et al., Cortical glutamate in migraine, Brain, 2017; 140(7):1859-1871
  • van de Bank BL, Emir UA, Boer VO, van Asten JJA, Wijnen JP, Kan HE, Oz G, Klomp DWJ and Scheenen TWJ. Multi-center reproducibility of neurochemical profiles in the human brain at 7 Tesla. NMR in Biomed. 2015 Mar;28(3):306-16.
  • van den Bogaard SJA, Dumas EM, Teeuwisse WM, Kan HE, Webb AG, van Buchem MA, Roos RAC, van der Grond J. Longitudinal metabolite changes in Huntington’s disease during disease onset. J. Huntington‘s disease, 2014;3(4):377-86 


  • Hermien E. Kan, h.e.kan@lumc.nl
  • Itamar Ronen, i.ronen@lumc.nl
  • Chloé Najac, c.f.najac@lumc.nl