The Mitochondrial Medicine Society

Advancing education, research, and global collaboration in clinical mitochondrial medicine

MRI & MRS in Mitochondrial Disease



The use of neuroimaging has not been universally accepted as criteria in the diagnostic evaluation of mitochondrial disease. Some schemes completely ignore nuclear magnetic resonance (MRI) findings while other diagnostic criteria include MRI findings . The ambiguity of using MRI as diagnostic criteria may be due to the heterogeneity of mitochondrial disease itself. Variation in clinical and biochemical findings vary between disease entities and even in patients having the same disease. For many mitochondrial disorders, mitochondrial DNA (mtDNA) and electron transport chain (ETC) diseases, morphological or structural changes of the brain are non-specific. Global MRI abnormalities occur in many patients. Both groups can display nonspecific delayed myelination pattern early in the course of the disease . Although MRI may be normal, especially in conditions of pure myopathy , most patients with central nervous system involvement have MRI abnormalities .


There are certain MRI findings highly sensitive and specific in diagnosis of mitochondrial disease. The most common specific MRI finding is a symmetrical signal abnormality of deep gray matter, high T2 and FLAIR and low T1 signal. Any deep structure can be involved and the character of the lesion can be either patchy or homogeneous. Leigh syndrome is the prototype disease, with imaging findings showing the diagnostic hallmark of Leigh syndrome. Denis Leigh described the neuropathological features of focal, bilateral, and symmetric necrotic lesions associated with demyelination, vascular proliferation and gliosis in the brain stem, diencephalon, basal ganglia, and cerebellum . The MRI high T2 signal represents prominent blood vessels and vascular proliferation in these deep gray structures. The volume of the lesion(s) is preserved on follow-up studies, in sharp contrast to the cerebral organic acidurias that lead to severe degeneration. MRI findings may differentiate distinct etiologies of Leigh syndrome as patients with Leigh syndrome due to SURF-1 mutations have lesions in the brain stem, subthalamic nuclei and possibly cerebellum with few patients having basal ganglial abnormalities . Patients with Leigh syndrome from other etiologies had T2 hyperintensities in the putamina with involvement of the caudate nuclei, globus pallidi, thalami, and brain stem with some patients with diffuse supratentorial white matter changes . Symmetric involvement of deep gray structures in the absence of hypoxia, ischemia, or infections is highly suspicious for a mitochondrial defect. Added justification for further work-up would be lactate peaks within the lesion on proton nuclear magnetic resonance spectroscopy.


Infarct-like, often transient lesions not confined to vascular territories are the imaging landmark of mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes, MELAS . Focal necrosis and laminar cortical necrotic changes with neuronal degeneration and mineral deposits within the basal ganglia are the histopathological correlates of this disease . On brain MRI, lesions are usually focal or multifocal involving signal changes in the grey and white matter predominantly in the occipital and parietal lobes . Lesions display T2 and FLAIR hyperintensities with a normal to slightly increased apparent diffusion coefficient . Acute ischemic stroke episodes show hyperintense T2 and FLAIR however, there is a significant reduction in apparent diffusion coefficient . Therefore diffusion-weighted sequences during the acute event can differentiate the two types of events are critical to the diagnostic work-up.


Kearns-Sayre/Pearson syndrome and chronic progressive external ophthalmoloplegia can express considerable overlap in signs and symptoms. The most common brain MRI findings are cerebral and cerebellar atrophy often with bilateral thalamic and basal ganglia (substantia nigra and globus pallidus) lesions . Although basal ganglia changes on MRI were not found in 7 patients with Kearns-Sayre syndrome in the series by Barragán-Campos et al (2005). The typical histopathological finding is status spongious, both gray and white matter are affected, most often the brain stem tegmentum and white matter of the cerebrum, cerebellum and basal ganglia . Involvement of the subcortical U fibers with sparing of the periventricular white matter differentiates Kearns-Sayre syndrome from other leukodystrophies, in which the “older” white matter is affected early in the course of the disease .


Most infants and children with mitochondrial disease do not have Leigh syndrome, MELAS, Kearns-Sayre/Pearson syndrome or chronic progressive external ophthalmoplegia but electron transport chain deficiencies. Although most electron transport chain abnormalities with clinical CNS involvement have MRI abnormalities, most are non-specific . However, when specific MRI findings are present in the appropriate clinical setting, it can guide the diagnostic approach. We feel that the presence of a deep gray matter signal abnormality with proper clinical signs and symptoms justifies further diagnostic work-up for a possible mitochondrial disease. An added modality of proton magnetic resonance imaging finding of a lactate peak would be further indication of a work-up. This non-invasive modality can help the multidisciplinary approach for diagnosis of a mitochondrial disorder.


Classification of the mitochondrial disorders is challenging, with the relationship between phenotype to genotype complex and often confusing. There is no single algorithm for the diagnosis; however in the setting of high clinical suspicion an impaired redox state shown within the CNS adds substantial weight to the diagnosis. Disruption of the respiratory chain and consequent depletion of NAD+ and NADP+ shifts the predominant catabolic metabolism from Krebs cycle to anerobic glycolysis. This produces an accumulation of pyruvate and its reduced product lactate. In patients who manifest characteristics of mitochondrial respiration disruption and in whom hypoxia, prolonged seizure and increased CNS demand for ATP can be excluded, an elevated lactate can be evidence that confirms the presence of a mitochondrial disorder.


Various MRS sampling programs are commercially available. Single or multivoxel sampling can be performed . The advantage to multivoxel sampling is that both gray and white matter can be sampled in a systematic fashion in various contiguous volumes through a particular slice of brain. While single voxel sampling varies depending on which regions the examiner picks, and therefore can vary in longitudinal studies and between patients.


Voxels can be sampled using one or multiple echo times; 35 msec (short), 144 msec (intermediate) and 288 msec (long). Spectra can be evaluated by comparing amplitudes of resonance signals; peak amplitudes are noted by assuming Lorentzian line shape and evaluating the baseline noise standard deviation. The lactate elevation is a 7 Hertz doublet at a chemical shift of 1.3 ppm, while NAA peak is found at 2.01 ppm and creatine/phosphocreatine at 3.05 ppm. Other important peaks are myo-inositol at 3.56 ppm and choline at 3.25 ppm. Using multiple echo times may help identify the 1.3 ppm lactate doublet peak from the background noise. On shorter TE times, 35 msec, the doublet is elevated while on intermediate TE times, 144 msec, the doublet is downgoing or inverted.


Proton magnetic resonance spectroscopy (MRS) can non-invasively measure regional and longitudinal metabolic information. The measurement of CSF and brain lactate has been shown to be helpful in the diagnosis and monitoring of mitochondrial disease . Studies have also shown that in disorders of electron transport chain function, low N-acetylaspartate/creatine (NAA/Cr) ratio was correlated with deep gray matter signal abnormality on magnetic resonance imaging (MRI) and lactate elevation. In one study, lactate peaks were noted in both gray and white matter in children with electron transport chain disorders .


There is a suggestion that MRS may be diagnostic of succinate dehydrogenase deficiency or complex II disorder. Brockmann et al. have described three cases of complex II dysfunction that presented with unclassificed leukoencephalopathy . All patients expressed a dramatic singlet peak at 2.40 ppm in cerebral and cerebellar white matter, which was not present in gray matter or the basal ganglia.


Bianchi et al. has described an abnormal chemical shift peak at 0.9 ppm that was found in both areas of abnormal lactate elevation as well as normal appearing areas in the cerebellum, white matter, and gray matter . This abnormality is thought to be branched chain amino acids and other lipids. It was not detected in normal controls. The authors noted the need to validate this funding in future studies but suggest that this peak may be of diagnostic value in mitochondrial diseases.


Several studies have shown that proton MRS abnormalities are usually only present in patients with CNS involvement and not seen in “pure” myopathy patients. The presence of lactate and decreased NAA/Cr ratio may depend on etiology of disease, i.e. electron transport chain dysfunction versus mtDNA mutation, or timing of scan to exacerbation of disease.


Like all diagnostic testing for possible mitochondrial disease, there is no one single test that accurately defines a patient with mitochondrial disease. Proton MRS is one diagnostic piece of the multifaceted testing that allows proper diagnosis. One study even suggests that proton MRS lactate elevation be included as one of the minor criteria for diagnosis.




Magnetic resonance spectroscopy (MRS) allows the non-invasive measurement of small molecules that are important in tissue biochemistry and metabolism (Danielson and Ross, 1999; Brandao and Comingues, 2004). This technique is based on the principles of nuclear magnetic resonance, the interaction between a molecule and an external magnetic field. The magnetic resonance imaging (MRI) scanner can produce specific magnetic fields and is used to generate both the MRI and MRS acquisition. This allows most clinical imaging centers to perform MRS. Furthermore, the lack of ionizing radiation makes it safe for repeated testing in infants and children.

The property of the molecule necessary for magnetic resonance is the nuclear spin, which is the property of proton, neutron, and atomic nuclei. The ability of a molecule to emit radio waves is related to a unique resonance frequency for each nuclei, the Larmor frequency. At the Larmor frequency the nuclei absorbs energy and this absorption causes the proton to alter its alignment or spin and can be detected in a magnetic field. Another type of molecular interaction that modifies the resonance of spin is coupling. Coupling occurs when there are multiple spins within a molecule; this interaction produces alterations in the local magnetic field around each nucleus. The resulting altered local magnetic field or coupling is independent of magnetic field strength. Both the Larmor frequency and coupling induce specific changes within nuclei that can be detected with MRS.

Nuclei of importance in biological processes include 1H, 31P, and 13C. The most commonly magnetically active nuclei studied in mitochondrial diseases are hydrogen (1H) and phosphorus (31P). However, carbon (13C) imaging is very useful in studying inborn errors of metabolism and flux through metabolic pathways and may become a more important tool in mitochondrial disease.

Each nuclei of a molecule has different physiochemical properties and limitations with respect to their MRS signal production. Parameters that affect the signal are field strength, natural abundance, typical concentration within the tissue, and spin state. In addition, to be detectable a molecule must be relatively mobile within the tissue. Most macromolecules with molecular weights greater than 20 kilodaltons do not have sufficient freedom of movement to be detected. Molecules of interest must also have sufficient concentration of atomic nuclei to generate enough signal to be detected above the background environmental noise examined. Generally, molecules need to be in the millimolar range for adequate detection.


The resultant output or signal of MRS is seen in various peaks (area under the peak) which correspond to the number of nuclei and hence the metabolite concentration. Depending on the types of nuclei and bonds within a molecule, the nuclei in a molecule experience different magnetic field strengths depending on their chemical environment. The variation in magnetic field strengths produces a specific resonance frequency called chemical shift, which is expressed as parts per million (ppm). Particular molecules have specific chemical shift and are relatively uniform in a particular tissue, such as the brain.


The quantification of spectral peaks plays an important role in MRS imaging. The peak area is proportional to the number of spins that produce the signal. Visual inspection is not reliable as molecules may have similar chemical shifts, differ slightly in the chemical shift and due to concentration variation may contaminant peak of interest, or be altered in behavior at different time to echo (TE) measurements. Most analyses currently use peak area ratios, the comparison of metabolites to one metabolite (i.e. creatine). This is because the acquired MRS signal is uncalibrated and given in machine units. Exact quantification of peak areas is currently an evolving science and when fully available will allow more precise comparison of values between patients or the same patient under different conditions.


The use of peak ratios (pattern analysis) to define “normal” versus “abnormal” has intrinsic problems as a paradigm to define particular spectra. There is an assumption that the denominator remains constant and is not a part of the disease process studied. Many MRS programs use creatine (3.03 ppm) as the denominator. Concentrations in the human brain are approximately 4.0 – 5.5 mmol/kgww for phosphocreatine and 4.8 – 5.6 mmole/kgww for creatine (Erecianska and Silver, 1989). Total creatine signal is reported higher in gray matter at 6.4 – 9.7 mM than in white matter at 5.2 – 5.7 mM (Pouwels and Frahm, 1998; Wang and Li, 1998). The creatine concentration is relatively stable with no changes reported with age or a variety of diseases (Saunders et al., 1999). These findings have lead to its common use a concentration reference. However, caution is needed as decreased levels are observed in tumors and stroke and increased levels with myotonic dystrophy (Chang et al., 1998). Furthermore, creatine is absence in disorders of creatine synthesis (Stockler et al., 1994) and increased during treatment (Stockler et al., 1996).

Proton MRS

Proton or 1H MRS is the most widely used spectroscopy technique used for investigations of metabolism within the brain. The advantage to proton MRS is that it shares the same radio-frequency range with routine MRI and is cost effective as most clinical MRI machines can be used for testing. Depending on field strength, approximately 80 or more brain metabolites can be distinquished (Ross, 2000). The most common chemicals studied are N-acetyl-L-aspartate (NAA), creatine (Cr), phosphocreatine (PCr), choline (Cho), myo-inositol (mI), lactate (Lac), glutamate (Glu), and glutamine (Gln). Some of these compounds are difficult to resolve at lower field strengths due to complex coupled spins, such as Glu and Gln which are found in a large shoulder peak at 2.05-2.5 ppm just adjacent to the NAA peak at 2.02 ppm (Kingsley et al., 2006).


The most common peaks studied are NAA (2.02 ppm), total Cr (including phosphocreatine, 3.03 ppm), Cho (3.22 ppm), mI (3.55 ppm), Lac (1.33 ppm), and succinate (2.39/2.40 ppm). Most of these peaks can be seen in Figure 1 with the exception of succinate which only becomes pronounced in complex II disorders within white matter and CSF (Brockman et al., 2002).


Metabolite peaks may vary according to the echo time (TE) used for investigation. Varying TE times can help separate metabolite of interest from metabolites having similar chemical shifts. For example, similar chemical shifts can be seen in Lac (1.33 ppm), propanediol (1.14 ppm), alanine (1.48 ppm) and underlying lipid (0.8 – 1.5 ppm). On short TE (TE 35) the Lac doublet peak will be up going while at intermediate TE (TE 144) the Lac doublet peak will be downward. Lipid may obscure the Lac signal at short TE times but at longer TE, the flip downward of Lac may clear up a previous non-descriptive spectrum. Propanediol is a solvent in the preparation of Phenobarbital and can contaminant the Lac signal if intervenous Phenobarbital is used at high concentrations. Another example, at longer TE, Glu, Gln and mI are not well visualized due to short relaxation times. The function of mI is not well understood but is thought to be essential for cell growth, acting as an osmolite, and a storage form for glucose (Ross, 1991). It has been proposed as an astrocyte marker (Brand et al., 1993). Altered levels have been associated with Alzheimer’s disease, hepatic encephalopathy, and brain injury (Shonk et al., 1995; Moats et al., 1994; Ross et al., 1998). At shorter TE, the mI peak is prominent and changes in concentration are better viewed.


On MRS, the Cho signal is a group of components, including free choline, phosphorylcholine, and phosphorylcholine, compromising brain myelin and fluid-cell membranes that resonate at 3.2 ppm (Tan et al., 1998; Miller et al., 1996). However, the complete metabolite profile contributing to the signal is unknown (Govindaraju et al., 2000). Elevations of Cho reflect membrane turnover and demyelination.


NAA has been shown to be predominately localized to neuron, axons, and dendrites (Simmons et al., 1991). Its function is not completely understood and is thought to act as an osmolite, a storage form of aspartate, a precursor of NAAG, and acetyl-CoA, as well as having a variety of other functions (Birken and Oldendorf, 1989). It has also been detected in other cell types such as oligodendrocyte precursors and astrocytes, suggesting that NAA may not be completely specific for neuronal processes (Urenjak et al., 1992; Bhakoo and Pearce, 2000). It is broken into aspirate and acetate by the enzyme asparto-acylase. NAA levels are found to be decreased in mitochondrial disease states (Bianchi et al., 2003). Overall, MRS measurements of NAA appear to be one of the best surrogate markers for neuronal integrity, however altered levels may not always represent neuron loss (Martin et al., 2001).


A component of the citric acid cycle, succinate is present in brain at low concentrations, 0.5 mmol/kgww (Klunk et al., 1996). In conventional MRS, succinate is present as a singlet at 2.39 ppm (Govindaraju et al., 2000) or 2.40 ppm (Brockman et al., 2002) and overlaps with resonances of glutamate and glutamine. Under normal conditions, the succinate peak is not visible on MRS spectrum. However, the level of succinate increases in succinate:ubiquinone oxidoreductase (E.C. deficiency and becomes visible on MRS (Brookmann et al., 2002).

Proton MRS and Mitochondrial Disease

Nuclear magnetic resonance imaging (MRI) is the mainstay of radiological evaluation of anatomic structures in neurometabolic diseases, in particular mitochondrial disease. However, many mitochondrial diseases have no MRI abnormalities or only nonspecific changes (Barkovich et al., 1995; van der Knaap et al., 1996; Valanne et al., 1998). Under conditions when metabolism shifts to anaerobic glycolysis, pyruvate and its reduced product lactate accumulate. Under such conditions of lactate accumulation and nonspecific MRI findings, MRS can be a valuable tool in the evaluation of suspected mitochondrial disease (Barkovich et al., 1993; Lin et al., 2003; Bianchi et al., 2003; Dinopoulos et al., 2005). Barkovich et al. (1993) detected Lac peaks in 5 of 5 patients with known Leigh syndrome and mitochondrial myopathy, encephalopathy, lactic acidosis and stroke-like episodes (MELAS). Nine out of 9 patients with MELAS were found to have Lac peaks in two other studies (Castillo et al., 1995; Wilichowski et al., 1999). There were 18 out of 21 patients with various respiratory chain deficiencies in another study (Saneto et al., 2001). However, just as serum lactate levels are only a moderately sensitive marker for mitochondrial disease, brain Lac peaks on MRS are also not 100% sensitive. Only 8 patients of 29 were found to have Lac peaks on MRS in one study (Lin et al., 2003) while only 2 of 11 patients in another study (Bianchi et al., 2003). Lac peaks have been shown to vary whether a patient is undergoing exacerbation of their disease (Bianchi et al., 2003) and if lesions are acute, subacute or chronic (Lin et al., 2003). Dectection of Lac signal may also depend on concentration as threshold of detection is between 0.5 and 1 mmol (Ross, 2000). Lack of Lac peak may also be due to differences in type of mitochondrial disease or location of the affected brain site. Other conditions can give rise to Lac peaks, such as seizures, stroke, and other metabolic conditions (Kingsley et al., 2006). The latter are usually readily excluded during the metabolic workup by clinical presentation, MRI, and biochemical investigations. In the scenario of clinical and biochemical suspicion for mitochondrial disease, the finding of a Lac peak would increase the possibility of the disease and further the workup plan.


Overall, MRS measurements of NAA appear to be one of the best surrogate markers for neuronal integrity. NAA is synthesized within the mitochondria in an energy dependent fashion (Clark, 1998). The suggestion is that NAA may represent a functional mitochondria marker within neuronal populations and reductions may reflect mitochondrial disease states. In 15 patients with mitochondrial disease syndromes, 93% showed NAA/Cr ratio reductions in cerebellum and 87% in cortical gray matter regions within some regions that appear normal on MRI (Bianchi et al., 2003). Interestingly, in this study, no decrease in the NAA/Cr signal was found in white matter. In another series, 16 patients with definitive mitochondrial disease consisting of known syndromes and respiratory chain deficiencies, 11 (69%) had decreased NAA/Cr ratio (Dinopoulos et al., 2005). In this study, both gray matter and white matter regions showed decreased NAA/Cr ratio. The NAA/Cr ratio signal changes were only seen in those patients with Lac peaks. In other studies, the apparent decrease in the NAA/Cr signal was reversible (Pavlakis et al., 1998; de Stefano et al., 1995). The decrease of NAA/Cr may vary depending on the type of disorder, the timing of investigation with respect to disease progression and region of investigation. As with Lac, NAA is non-specific with other disorders presenting with either increase or decrease the NAA/Cr ratio (Kingsley et al., 2006; Martin et al., 2001). However, usually the clinical and biochemical work-up will segregate these disorders from primary mitochondrial disease.


Cho signal changes have not been well reported. Whether this represents the lack of change or heterogeneous change that small sample populations may have not appreciated is not known. One study demonstrated reductions in the Cho/Cr ratio in 40% of patients with reduced NAA/Cr and 53% in patients with reduced NAA/Cr ratio and Lac (Bianchi et al., 2003).


Complex II (succinate:ubiquinone oxidoreductase, E.C. catalyzes the oxidation of succinate to fumarate in the Krebs cycle and carries electrons to the ubiquinone pool of the respiratory chain (Ackrell et al., 1992). When enzyme activity is deficient there are significant increases in succinate concentrations. In 3 patients with a complex II deficiency, a very large signal was noted at 2.4 ppm within white matter. There were also reductions in NAA/Cr with increased Lac signal. In contrast, within cortical gray matter, only decreased NAA/Cr ratio was observed (Brockmann et al., 2002). Two interesting points are noted from this study, one is that multivoxel acquisition or at least using single voxels over both gray and white matter are essential to detect this respiratory chain disease. Second, is the possible high specificity of the enlarged 2.4 ppm peak of succinate in complex II deficient disease.


Proton MRS represents a useful tool in the non-invasive investigation of neurometabolic disorders, in particular mitochondrial disease. It supplies additional information to conventional MRI as metabolic changes can be visualized even in areas of brain that appear normal. It is non-invasive and does not require radiation exposure, so it can be used to follow disease conditions over time. It uses conventional MRI acquisition and therefore is available to most neuroimaging centers. In addition, it does not need extraordinary extra time to perform and can be added to routine MRI imaging studies (Lin et al., 2005).


MRS findings are a reflection of the clinical heterogeneity in mitochondrial disease. Appropriate testing during particular stages of disease as well as examining the proper brain area(s) will increase sensitivity of disease detection. Studies using larger numbers of patients with defined diseases of particular genetic etiologies will enhance the diagnostic value of MRS results. So although the present sensitivity of abnormal findings in the determination of a particular mitochondrial disease is not extremely high, data derived from MRS make it a useful part of the investigation rubric of mitochondrial disease.



1. Danielson ER, Ross B. Magnetic Resonance Spectroscopy Diagnosis of Neurological Disease. Marcel Kekker: New York, 1999.

2. Brandao LA, Comingues RC. MR Spectroscopy of the Brain. Lippincott Williams and Wilkens: Philadelphia, 2004.

3. Ross Bd. Real or imaginary? Human metabolism through nuclear magnetism IUBMB Life 2000; 50:177-187.

4. Kingsley PB, Shah TC, Woldenberg R. Identification of diffuse and focal brain lesions by clinical magnetic resonance spectroscopy. NMR Biomed 2006;19:435-462.

5. Simmons MI, Frondoza CG, Coyle JT. Immunocytochemical localization of N-acetyl-aspartate with monoclonal antibodies. Neuroscience 1991;45:37-45.

6. Birken DL, Oldendorf WH. N-Acetyl-L-aspartic acid: a literature review of a compound prominent in 1H-NMR spectroscopic studies of brain. Neurosci Biobehav Rev 1989;13:23-31.

7. Urenjak J, Williams SR, Gadian DG, Noble M. Specific expression of N-acetylaspartate in neurons, oligodendrocyte-type-2 astrocytes prgenitors, and immature oligdendrocytes in vitro. J Neurochem 1992;59:55-61.

8. Bhakoo KK, Pearce D. In vitro expressioin of N-acetyl aspartate by oligdendroyctes: implications for proton magnetic resonance spectroscopy signal in vivo. J Neurochem 2000;74:254-262.

9. Bianchi MC, Tosetti M, Battini R, Manca ML, Mancuso M, Gioni G, Canapicchi R, Siciliano. Proton MR spectroscopy of mitochondrial diseases: analysis of brain metabolic abnormalities and their possible diagnostic relevance. AJNR Am J Neuroradiol 2003;24:1958-1996.

10. Dinopoulos A, Cecil KM, Schapiro MB, Papdimitriou A, hadjigeorgiou GM, Wong B, deGrauw T, Egelhoff JC. Brain MRI and proton MRS findings in infants and children with respiratory chain defects. Neuropediatrics 2005;36:290-301.

11. Martin E, Capone A, Schneider J, Hennig J, Thiel T. Absence of N-acetylaspartate in the human brain: impact on neurospectroscopy? Ann Neurol 2001;49:518-521.

12. Erecianska M, Silver IA. ATP and brain function. J Cereb Blood Flow Metab 1989;9:2-19.

13. Pouwels PJ, Frahm J. Regional metabolite concentrations in human brain as determined by quantitative localized proton MRS. Magn Reson Med 1998;39:53-60.

14. Wang Y, Li SJ. Differentiation of metabolic concentrations between gray matter and white matter of human brain by in vivo 1H magnetic resonance spectroscopy. Magn Reson Med 1998;39:28-33.

15. Saunders DE, Howe FA, va den Boogaart A, Griffiths JR, Brown MM. Aging of the adult human brain: In vivo quantitation of metabolite content with proton magnetic resonance spectroscopy. J Magn Reson Imaging 1999;9:711-716.

16. Chang L, Ernst T, Osborn D, Seltzer W, Leonido-Yee M, Poland RE. Proton spectroscopy in myotonic dystrophy: correlations with CTG repeats [See comments] Arch Neurol 1998;55:305-311.

17. Stockler S. Holzback U. Hanefeld F, Marquardt I, Helms G, Requart M, Hanicke W, Frahm J. Creatine deficiency in the brain: a new, treatable inborn error of metabolism. Pediatr Res 1994;36:409-413.

18. Stockler S, Hanefeld F, Frahm J. Creatine replacement therapy in guanidinoacetate methyltransferase deficiency, a novel inborn error of metabolism. Lancet 1996;348:789-790.

19. Brockmann K, Bjornstad A, Dechent P, Korenke G, Smeitink J, Frans Trijbels JM, Sthanassopoulos S, Villagran R, Skjeldal OH, Wilchowski E, Frahm J, Hanefeld F. Succinate in dystrophic white matter: a proton magnetic resonance spectroscopy finding characteristic for complex II deficiency. Ann Neurol 2002;52:38-46.

20. Ross BD. Bichemical consideration in 1H spectroscopy, glutamate and glutamine; myoinositol and related metabolites. NMR Biomed 1991;4:59-63.

21. Brand A, Richter-Landsberg C, Leibfritz D. Multnuclear NMR studies on the energy metabolism of glial and neuronal cells. Devel Neurosci 1993;15:289-298.

22. Shonk TK, Moats RA, Gifford P, Michaelis T, Mandigo JC, Izumi J, Ross BD. Probable Alzeheimer disease: diagnosis with proton MR spectroscopy. Radiology 1995;195:65-72.

23. Moats RA, Ernst T, Shonk TK, Ross BD. Abnormal cerebral metabolite concentrations in patients with probable Alzeheimer disease. Magn Reson Med 1994;32:110-115.

24. Ross BD, Ernst T, Kreis R, Haseler IJ, Bayer S, Danielsen E, Bluml S, Shonk T, Mandigo JC, Caton W, Clark C, Jensen SW, Lehman NL, Arcinue E, Pudenz R, Shelden CH. 1H MRS in acute traumatic brain injury. J Magn Reson Imag 1998;8:829-840.

25. Klunk WE, Xu C, Panchalingam K, McClure RJ, Pettegrew JW. Quantitative 1H and 31P MRS of PCA extracts of postmortem Alzheimer’s disease brain. Neurobiol Aging 1996;17:349-357.

26. Ackrell BAC, Johnson MK, Gunsalus RP, Cecchini G. Structure and function of sucinate dehydrogenase and fumarate reductase. In Muller F, ed. Biochemistry of flavoenzymes. Boca Raton FL: CRC, 1992:229-297.

27. Barkovich AJ, Good WV, Koch TK, Berg BO. Mitochondrial disorders: analysis of their clinical and imaging characteristics. AJNR Am J Neuroradiol 1993;14:1119-1137.

28. van der Knapp MS, Jakobs C. Valk J. Magnetic resonance imaging in lactic acidosis. J Inherit Metab Dis 1996;19:535-547.

29. Valanne L, Ketonen L, Majander A, Suomalainen A, Pihko H. Neuroradiologic findings in children with mitochondrial disorders. AJNR Am J Neuroradiol 1998;19:369-377.

30. Tan J, Bluml S, Hoany T, Dubowitz D, Mevenkamp G, Ross B. Lack of effect of oral choline supplement on the concentrations of choline metabolites in human brain. Magn Reson Med 1998;39:1005-1010.

31. Miller BL, Chang L, Booth R, Ernst T, Cornford M, Nikas D, McBride D, Jenden DJ. In vivo 1H MRS choline: correlation with in vitro chemistry/histology. Life Sci 1996;58:1929-1935.

32. Govindaragju V, Young K, Maudsley AA. Proton NMR chemical shifts and coupling constants for brain metabolites. NMR in Biomed 2000;13:129-153.

33. Castilo M, Kwock L, Green C. MELAS syndrome: imaging and proton MR spectroscopic findings. AJNR Am J Neuroradiol 1995;16:233-239.

34. Wilichowski E, Pouwels PJ, Frahm J, Hanefeld F. Quantitative proton magnetic resonance spectroscopy of cerebral metabolic disturbances in patients with MELAS. Neuropediatrics 1999;30:256-263.

35. Ross BD. A biochemistry primer for neuroradiologist. In: Advanced imaging symposium: Preparing the neuroradiologist for the new millennium. Oak Brook IL: American Society of Neuroradioogy 2000;13-27.

36. Saneto RP, Cohen BH, Ruggieri P. Hoppel CL. MRS detection of CNS lactate peaks in primary mitochondrial cytopathies. Neurology 2001;56:A42 (suppl 3).

37. Lin DDM, Crawford TO, Barker PB. Proton MR spectroscopy in the diagnostic evaluation of suspected mitochondrial disease. AJNR Am J Neuroradiol 2003;24:33-41.

38. Matalon R, Kaul R, Casanova J, Michals K, Johnson A, Rapin I, Gashkoff P. Deanching M. SSIEM Award. Aspartoacylase deficiency: the enzyme defect in Canavan disease. J Inherit Metab Dis 1989;12(Suppl 2):329-331.

39. Clark JB. N-acetyl aspartate: a marker for neuronal loss or mitochondrial dysfunction. Develop Neurosci 1998;20:271-276.

40. Pavlakis SG, Kingsley PB, Kaplan GP, Stacpoole PW, O’Shea M, Lustbader D. Magnetic resonance spectroscopy: use in monitoring MELAS treatment. Arch Neurol 1998;55:849-852.

41. de Stefano N, Matthews PM, Arnold DL. Reversibe decreses in N-acetylaspartate after acute brain injury. Magn Reson Med 1995:34:721-727.

42. Ackrell BAC, Johnson MK, Gunsalus RP, Cecchini G. Structural and function of succinate dehydrogenase and fumarate reductase. In: Muller F, ed. Biochemistry of flavoenzymes. Boca Raton, FL: CRC, 1992:229-297.

43. Lin A, Ross BD, Harris K, Wong W. Efficacy of proton magnetic resonance spectroscopy in neurological diagnosis and neurtherapeutic decision making. NeuroRx 2005;2:197-214.