DR. NGOC

1. BRAIN MR SPECTROSCOPY
Radiologist Le Thi Kim Ngoc

2. Introduction of MRS
 Magnetic Resonance Spectroscopy (MRS) :an
analytical method in chemistry that enables the
identification and quantification of metabolites in
samples.
 Differs from conventional Magnetic Resonance
Imaging (MRI): spectra provide physiological and
chemical information instead of anatomy

3. Introduction
 MR spectra may be obtained from different nuclei :1
H
(proton), 23
Na (sodium), 31
P (phosphorus). The
most common nuclei : 1
H (proton) (easier to perform
and higher signal-to-noise)
 Proton MRS : within 10-15 minutes, + conventional MR
imaging protocols.
 Monitor biochemical changes in neurologic disease
 Interpretation : be correlated with the MR images before
making a final diagnosis

4. Physical Basis
 Conventional MRI: 0.2 to 3T.
 MRS: to detect weak signals from metabolites, a
higher strength field is required (1.5T or more).
 Higher field strength units have the advantage:
higher signal-to-noise ratio (SNR), better
resolution and shorter acquisition times

5. BASIC PHYSICAL
PRINCIPLES
 H-MRS: the chemical shift properties of the atom
 When a tissue is exposed to an external magnetic
field, its nuclei will resonate at a frequency (f) that is
given by the Larmor equation
f = γBo
 gyromagnetic ratio (γ) is a constant of each
nuclear species
 the external magnetic field (Bo) and the local
microenvironment.

6. BASIC PHYSICAL
PRINCIPLES
 interactions of nuclei with the surrounding
molecules - change in the local magnetic field -
change on the spin frequency of the atom (a
phenomenon called chemical shift)
 The value of this difference in RF gives information
about the molecular group carrying H and is
expressed in parts per million (ppm).

7. The MR spectrum
 the x axis-
metabolite
frequency in
ppm according
to the chemical
shift
 the y axis that
corresponds to
the peak
amplitude

8. BASIC PHYSICAL PRINCIPLES
 The resonant frequencies of protons : 10 MHz (0.3
T) - 300 MHz (7 T)
 At 1.5 T, the metabolites :63,000,000 and
64,000,000 Hertz- parts per million (ppm). NAA at
2.0 ppm and other metabolites fall into their proper
positions on the spectral line the ppm scale :it reads
from right to left

9. Techniques
 Anatomical images-
select a volume of
interest (VOI)-
spectrum will be
acquired
 different techniques:
single- and multi-voxel
imaging using both
long and short echo
times (TE)

10. Single-Voxel Spectroscopy
 signal is obtained from a voxel
previously selected, in 3D
 Two techniques for acquisition
of SVS H-MRS spectra:
pointed-resolved spectroscopy
(PRESS) and stimulated echo
acquisition mode (STEAM).
 PRESS - mostly used SVS
technique : doubles SNR, better
spectral quality, more artifact
 STEAM - shorter than PRESS
(short TE and precise volume
selection is needed), lower SNR
than PRESS.

11. Multivoxel technique
 obtain simultaneously many
voxels and a spatial
distribution of the metabolites
within a single sequence
 Instead of the anatomical
information ( conventional
MRI), the MRS signal results in
a spectrum of metabolites with
different frequencies
 The same sequences used for
SVS (STEAM or PRESS)
 1D, 2Dor 3D to sample the k-
space

12. SVS vs MRSI
 SVS : high quality spectrum, a short scan time, and good
field homogeneity, usually obtained with short TE since
longer TE has decreased signal due to T2 relaxation.
SVS is used to obtain an accurate quantification of the
metabolites but spectrum in a limited brain region
 MRSI is spatial distribution but the quantification of the
metabolites is not as precise when using MRSI technique
because of voxel bleeding. Therefore,MRSI can be used
to determinate spatial heterogeneity.

13. Short TE vs long TE
 MRS with different TEs that result in distinct spectra
 Short TE: 20 - 40 ms, higher SNR and less signal loss,
more metabolies peaks, such as myoinositol and
glutamine-glutamate ( not detected with long TE) .Since
more peaks are shown on the spectrum, overlap in
quantifying the peaks of metabolites.
 Long TEs, from 135 to 288 ms: worse SNR, more
simple spectra due to suppression of some signals.
Thus, the spectra are less noisy but have a limited
number of sharp resonances.

14. Short TE vs long TE
Fig. 8. Spectrum obtained with TE = 30ms (A) and TE = 135ms (B). Note the inverted lactate
peak (doublet) with long TE acquisition and the more number of sharps resonance with short TE.
Cho– choline; Cr- creatine NAA– N-acetylaspartate; Ins dd1– myoinositol.

15. Technique
 The single voxel, short TE, to make the initial diagnosis, because
the signal-to-noise is high and all metabolites are represented.
 Multi-voxel, long TE, to further characterize different regions of a
mass and to assess brain parenchyma around or adjacent to the
mass. Multi-voxel, long TE techniques are also used to assess
response to therapy and to search for tumor recurrence

16. Water Suppression
 MRS-visible brain metabolites : low concentration in brain
tissues.
 Water : the most abundant-spectrum is much higher than
that of other metabolites (the signal of water is 100.000
times greater than that of other metabolites)
 To avoid high peak from water to be superimpose on the
signal of other brain metabolites- water suppression
techniques
 The most commonly used technique is chemical shift
selective water suppression (CHESS) which pre-saturates
water signal
 Other techniques : Variable Pulse power and Optimized
Relaxation Delays (VAPOR) and Water suppression
Enhanced Through T1 effects (WET).

17. Fig. 9. Water signal suppressing with CHESS. Spectrum before CHESS
(A) and after CHESS (B). CHESS reduces signal from water by a factor of
1000 allowing brain metabolites to be depicted on the spectrum.

18. Artifacts
 MRS is prone to artifacts: Motion, poor water or lipid
suppressions, field inhomogeneity, eddy currents, and
chemical shift displacement ..
 One of the most important factors: the homogeneity of
the magnetic field. Poor field homogeneity : a lower SNR
and broadening of the width of the peaks.
 For brain MRS, some regions are more susceptible to
this artifact: near bone structures and air tissue-
interfaces. Therefore placement of the VOI should be
avoided near areas such as anterior temporal and frontal
lobes. Paramagnetic devices

19. Higher Fields H-MRS
 Higher field MRI (3T, 7T and above): better SNR and
faster acquisitions factor which are important in sick
patients that cannot hold still
 H-MRS is more sensitive to magnetic field
inhomogeneity: artifacts due to eddy currents and
Chemical shift displacement
 Receiver coils: The use of multiple radiofrequency:
higher local sensitivity and results in higher SNR.
These coils also allow a more extended coverage of
the brain.

20. Spectra
 H-MRS allows the detection of brain metabolites.
 The metabolite changes: structural abnormalities.
MRS can demonstrate abnormalities before MRI does
 Spectral variations according to the technique, patient
age, and brain region.

21. Spectra
 1H spectra of metabolites are
shown on x and y axes.
 The x, horizontal, axis: the
chemical shift of the
metabolites (ppm) increases
from R to L
 The y, vertical, axis: signal
amplitude of the metabolites.
The height of metabolic peak
refers to a relative
concentration and the area
under the curve to metabolite
concentration

22. Main brain metabolites:
N-acetyl aspartate (NAA)
 Peak of NAA is the highest
peak in normal brain.
 N-acetyl aspartate : 2.02
ppm
 NAA is exclusively found in
the nervous system in both
grey and white matter. It is
a marker of neuronal and
axonal viability and density.

23. Main brain metabolites: NAA
 Absence or decreased concentration of NAA is a
sign of neuronal loss or degradation.
 Neuronal destruction from malignant neoplasms and
many white matter diseases result in decreased
concentration of NAA.
 In contrast, increased NAA is nearly specific for
Canavan disease. NAA is not demonstrated in extra-
axial lesions such as meningiomas or intra-axial
ones originating from outside of the brain such as
metastases.

24. Main brain metabolites: Creatine (Cr)
 The peak of Cr spectrum is at
3.02 ppm..
 Cr is a marker of energetic
systems and intracellular
metabolism.
 Concentration of Cr is
relatively constant, the most
stable cerebral metabolite - as
an internal reference for
calculating metabolite ratios.
 However, there are regional
and individual variability in Cr
concentrations.

25. Main brain metabolites: Cr
 In brain tumors: reduced Cr. On the other hand,
gliosis may cause minimally increased Cr due
to increased density of glial cells (glial
proliferation).

26. Main metabolites: Choline
 Its peak is assigned at
3.22 ppm
 represents the sum of
choline and choline-
containing compounds
 Cho is a marker of
cellular membrane
turnover (phospholipids
synthesis and
degradation) reflecting
cellular proliferation

27. Main metabolites: Cho
 Cho = Tumor marker or cell
 Active tumor growth :increase in Cho, since there is
above-normal production of cells, persistent Cho
elevation. In tumors, Cho levels correlate with
degree of malignancy reflecting of cellularity
 Other processes can release or increase Cho
besides tumor; multiple sclerosis or infarctions.
(from gliosis or ischemic damage to myelin) or
inflammation (glial proliferation). Hence elevated
Cho is nonspecific. This can be a transient effect.

28. Main metabolites: Lactate
 Lactate (lactic acid) is seen
as a doublet (two peaks
close to one another) at
1.33 ppm and is a by-
product of anaerobic
metabolism.
 Lipids resonate at the 0.9
to 1.2 ppm range.
 Both are released with cell
destruction or synthesized
in necrosis.

29. Main metabolites: Lactate
 Lip and Lac peaks are absent under normal conditions
 Lactate and lipid : present in aggressive disease
processes.
 Lac is a product of anaerobic glycolysis - anaerobic
metabolism : cerebral hypoxia, ischemia, seizures and
metabolic disorders (especially mitochondrial ones).
 Increased Lac signals also occur with macrophage
accumulation (e.g. acute inflammation).
 Lac also accumulates in tissues with poor washout such
as cysts, in abscess, normal pressure hydrocephalus,
and necrotic and cystic tumors

30. Main metabolites: Lipid
 Lipids: cell membranes, not
visualized on long TE (very short
relaxation time)
 There are two peaks of lipids: at
1.3 ppm and 0.9 ppm
 Absent in the normal brain,
(lipids may result from improper
voxel selection- adjacent fatty
tissues)
 Lipid peak: cellular membrane
breakdown or necrosis such as
in metastases or primary
malignant tumors.

31. Main metabolites:
Myoinositol (Myo)
 Myo: is a simple sugar, at
3.56 ppm. Myo is
considered a glial marker
(primarily synthesized in
glial cells)
 Myo may represent a
product of myelin
degradation.
 Elevated Myo: proliferation
of glial cells in
inflammation, in gliosis,
astrocytosis and in
Alzheimer’s disease

32. Main metabolites:
Alanine (Ala)
 Ala is an amino acid, doublet
centered at 1.48 ppm. This
peak is located above the
baseline in spectra obtained
with short/long TE and inverts
below the baseline on with
TE= 135-144 msec .
 Its peak may be obscured by
Lac (at1.33 ppm).
 The function of Ala is uncertain
 Increased concentration of Ala
: oxidative metabolism defects.
In tumors: elevated level of Ala
is specific for meningiomas.

33. Main metabolites:
Glutamate-Glutamine (Glx)
 Glx : complex peaks =
glutamate (Glu) + Glutamine
(Gln) + gamma-aminobutyric
acid (GABA), assigned at
2.05-2.50 ppm.
 These metabolite peaks are
difficult to separate at 1.5 T.
 Glu is an important excitatory
neurotransmitter
 Elevated concentration of Gln
is found in a few diseases
such as hepatic
encephalopathy

34. Obsevable Proton Metabolites
ppm Metabolite Properties
0,9-1,4 Lipid Product of brain destruction
1,3 Lactat Product of an aerobic glycolysis
2,0 NAA Neuronal maker
2,2-2,4 Glutamine/GAB
A
Neurotransmitter
3,0 Creatine Energy Metabolism
3,2 Choline Cell membrane maker
3,5 Myo-inositol Glial cell maker, osmolyte hormone receptor
mechanisim
1,2 Ethanol Triplet
1,48 Alanine Present in meningioma
3,4&3,8 Glucose Increased in diabete
3,8 Manitol Rx for increased ICP

35. Metabolite ratios
Normal Abnormal
NAA/Cr 2 <1,6
NAA/Cho 1,6 <1,2
Cho/Cr 1,2 >1,5

37. Regional variations of the
spectra
 Metabolites may varies from brain regions:
white and grey matters, supra and infratentorial
 The Cho concentrations are higher in white
matter than in gray matter but Cr is higher in
grey matter
 NAA are not significantly different in W and G
 The cerebellar levels of Cho are higher than
the supratentorial levels

38. Developmental variations.
Spectra in pediatrics
 Newborn : low NAA, high Cho, and high Myo levels on
MR spectroscopy
 As age ↑ , NAA and choline-containing compounds and
Myo become ↓. NAA reflects brain maturation ,
correlates with myelination
 These metabolites gradually approach the adult pattern
by 1 to 2 year, is practically constant by 4 years of age
 Creatine and phosphocreatine are constant - reference
values
 A small amount of Lac may be seen in newborn brains

40. Spectra in elderly
 With aging, the Cho concentration in
gray matter is significantly increased.
 Ongoing reseach

41. Clinical Application
 Brain Tumors
 Radiation injury
 Human Immunodeficiency Virus
(HIV)Infection
 Degenerative Disorders of the
Elderly:Alzheimer and Parkinson Diseases
 Inborn Error of Metabolism
 Hepatic Encephalopathy
 Cerebral Ischemia

42. Some Innovative Applications of MR
Spectroscopy
 Measurement of Psychoactive Drugs
 Neurofibromatosis Type 1
 Cerebral Heterotopias
 Multiple Sclerosis

43. MR spectroscopy of neoplasm
 Brain tumors are currently the main application
of H-MRS, used as a complement to
conventional MRI, along with other advanced
techniques
 Brain tumors : confirming the diagnosis, grading
the malignancy, and distinguishing radiation
necrosis from residual/recurrent neoplasm

44. MRS in brain tumor: TE and Voxel
 The most relevant parameter : TE
- Short TE : more peaks than long TE, differential diagnosis of
brain masses and for grading tumors. Myo is a marker for low
grade gliomas, only seen on short TE .
- Longer TEs: a limited number of peaks making it easier to
analyze.
- Long TEs varying from 135-140ms also invert peaks of Lac
and Ala. This inversion is important for differentiating between
these peaks and lipids since they commonly overlap
 The accuracy of MRS is greatest in voxels at the enhancing
edge of a lesion, avoiding areas of necrosis, hemorrhage,
calcification, or cysts.


46. MRS in brain tumor: MRSI and SVS
 MRSI is preferable to SVS : spatial distribution, a
spectrum of a lesion and the adjacent tissues,
tumor heterogeneity. However, MRSI is generally
combined with long TE
 SVS is faster, using both long and short TEs. The
VOI should be placed within the mass, avoiding
contamination from adjacent tissues. An identical
VOI must be positioned on the homologous region
of the contralateral hemisphere for comparison

48. MRS in brain tumor
 The typical H-
MRS spectrum
for a brain tumor
is one of high
level of Cho, low
NAA and minor
changes in Cr

49. MRS in brain tumor: Cho
 Elevation of Cho is seen in all neoplastic lesions.
 Cho increase : cellular membrane turnover which
reflects cellular proliferation, correlated with cell density.
 Cho peak is usually higher in the center of a solid
neoplastic mass and decreases peripherally. Cho signal
is consistently low in necrotic areas
 Diagnosis and progression of tumor, treatment
response. The degree of elevation of Cho correlates
with the histologic grade of malignancy and is helpful in
distinguishing tumors from non-neoplastic disease
processes.

51. MRS in brain tumor: NAA and Cr
 Decrease NAA. This metabolite is a neuronal
marker : destruction and displacement of
normal tissue.
 Absence of NAA in an intra-axial tumor
generally implies an origin outside of the
central nervous system (metastasis) or a highly
malignant tumor that has destroyed all neurons
in that location.
 Cr : slightly variable in brain tumors. It changes
according to tumor type and grade.

53. MRS in brain tumor
 Cho elevation is usually evidencedby
increase in Cho/NAA or Cho/Cr ratios,
rather than its absolute concentration.
 Absolute Cho concentration is
susceptible to many errors
 Therefore, Cho/NAA and Cho/Cr ratios are
accurate for establishing Cho levels in
brain neoplasms.

54. MRS in differentiaton brain
tumor and others
 A low grade glioma (LGG) from stroke or focal cortical dysplasia:
increased levels of Cho
 A giant demyelinating plaque usually shows high Cho and low NAA
levels, increase Lac. Cho increase: transient
 Brain abscess and neoplasms:
+ restricted diffusion
+ VOI is positioned in the enhancing area - presence of Cho favors a
neoplasm.
+ VOI is positioned in the cystic area of a lesion: Presence of acetate,
succinate, and amino acids (AAs) such as valine, alanine, and
leucine in the core of the lesion have high sensitivity for pyogenic
abscess. These peaks are not seen in tumors (pyogenic brain
abscess that are under antibiotic therapy these peaks may be
absent)

56. MRS in Glioma
 Gliomas: the most common neuroepithelial tumors.
They originate from glial cells (e.g. astrocytes or
oligodentrocytes)
 Astrocytomas : low grade (grade I and II, benign) and
high grade (grade III - anaplastic gliomas and IV
glioblastoma multiforme - malignant).
 A typical astrocytoma: ↓NAA, ↓ Cr, and ↑Cho levels,
helpful in distinguishing tumors from non-neoplastic
disease processes

57. MRS in Glioma
 Lip and Lac peaks are absent under normal
conditions.
 Lipid peak indicates necrosis in malignant tumors.
 Lac, a product of anaerobic glucolysis and
accumulates in necrotic portions of tumors.
 Presence of Lip and Lac correlate with necrosis in
high grade gliomas.An elevated lactate level is
frequently found in high-grade malignancies

58. MRI in grading brain tumors
 Accurate grading of gliomas on the basis of MRS alone
may be difficult:
 MRS + conventional MR+ other advanced MRI techniques
(perfusion MRI..) = precise grading.
- Conventional MRI: contrast enhancement, surrounding
edema, signal heterogeneity, necrosis, hemorrhage and
midline crossing
- Perfusion MRI (high rCBV) suggest a high grade.
- MRS : Cho, NAA, Cr, Ratio, Myo, Lac, Lip

59. MRS in grading brain tumors
 High grade gliomas demonstrate marked elevation of Cho,
decreased NAA
 The degree of elevation of Cho correlates with the histologic
grade of malignancy and is. Cho/Cr, Cho/NAA ↑, threshold
values of metabolite ratios for grading of gliomas are not well
established. Cho/Cr is the most frequently used ratio: cutoff
value of 2.5
 Presence of Lac and Lip.
 Myo is high in low grade gliomas and decreases with
increasing grades of tumors. Low grade gliomas : ↑ Myo
levels compared with high grade gliomas. This may be due to
low mitotic index in low grade gliomas and, thus, lower
mitogens (substances that trigger cell mitosis), results in Myo
accumulation, characteristic of grade II astrocytomas.

63. Secondary (Metastatic) Neoplasms
 Typical MRS of secondary neoplasms: ↑ lipid, lactate,
and choline and reduced or absent NAA
 Distinguishing metastases from high-grade primary
neoplasm: Primary neoplasms (infiltrate surrounding
brain tissue). Interrogation of areas outside the
enhancing portion of the lesion (the high signal
intensity on T2 weighted imaging seen in the
perilesional area) demonstrates ↑ Cho/Cr ratio, ↑
Cho/NAA ratio only in high grade gliomas.This feature
is consistent with the pathological findings of infiltrating
tumor cells in areas of edema not seen in metastases.
( In one study, a choline/NAA ratio of greater than 1
had an accuracy of 100%)

65. Lymphoma
 Typical MRS of lymphoma: ↑ lipid, lactate, and
choline and ↓ NAA signal
 MRS of lymphoma in AIDS patients: mild to
moderately ↑ lactate and lipid signals, along with a
prominent choline peak and ↓ NAA, creatine, and
myoinositol signals. This pattern can help in
differentiating lymphoma from toxoplasmosis, which
typically ↑ lactate and lipid signals but absence of
the other metabolites in MR spectra

66. Tumefactive Demyelinating
Lesions
 Typical MRS for tumefactive demyelinating
lesions : ↑ choline peak and ↓ NAA signal ,
presence of lactate (difficult to distinguish a
tumefactive demyelinating lesion from neoplastic
lesions at MR spectroscopy)
 In multiple sclerosis, the spectroscopic
abnormalities are not limited to visible lesions,
since normal-appearing areas of white matter were
shown to have reduced NAA signal (compared with
NAA signals in healthy control subjects). In the
early stages of the disease, an increased
myoinositol peak may be more apparent in normal-
appearing white matter than reduced NAA signal

68. Encephalitis
 Typical MRS features for encephalitis: ↑ lactate,
choline, and myoinositol and ↓ NAA signal
 Encephalitis resemble low-grade gliomas, with
reduction of the NAA signal and elevation of the
choline and myoinositol peaks A lactate peak is an
inconsistent finding
 After the initial acute phase of encephalitis, gradual
normalization of the MR spectrum in about 1 year

69. Brain Abscesses
 Typical MRS of brain abscesses :elevated peaks of amino
acid, lactate, alanine, acetate, pyruvate, and succinate
and absent signals of NAA, creatine, and choline
 Abscesses have a distinct spectroscopic pattern that
allows differentiation from other entities.
 - elevation of choline and absence of signal from a
variety of amino acids, acetate, and succinate :neoplastic
process,
 - whereas the other peaks listed above—alanine, acetate,
pyruvate, and succinate—favor abscesses
 Tuberculous abscesses typically have high lipid and
lactate peaks. These abscesses have no peaks for amino
acids (leucine, isoleucine, and valine) at 0.9 ppm,
succinate at 2.41 ppm, acetate at 1.92 ppm, and alanine
at 1.48 ppm, in contrast to pyogenic abscesses, which
have peaks for all these metabolites

71. Radiation necrosis
 A common clinical problem: distinguishing tumor
recurrence from radiation effects ( following surgery
and radiation therapy).
 ↑ Choline is a marker for recurrent tumor.
 Radiation change : ↓ NAA, ↓ Creatine, and ↓
Choline. If radiation necrosis is present: ↑ lipids and
lactate.

73. MRS : Cerebral Ischemia and
Infarction
 When the brain becomes ischemic - anaerobic
glycolysis and lactate accumulates.
 ↑↑ lactate is the key spectroscopic feature of
cerebral hypoxia and ischemia. Choline ↑ , and
NAA and creatine ↓ .
 If cerebral infarction : lipids ↑

75. Inborn Error of Metabolism
 The diagnosis of an inborn error of metabolism: challenging and
mainly based on clinical and laboratorial findings,and genetic
tests.
 Brain MRI : narrowing the differential diagnosis, establishing a
final diagnosis.
 Since these disorders are caused by inherited enzymatic
defects, concentrations of some metabolites may be abnormally
low or high. Metabolites with a very small concentration in brain
tissue are not depicted on H-MRS. In these cases, the
spectrum changes usually correspond to a general pathology,
such as demyelination or ischemia.
 On some diseases: H-MRS may identify a specific biomarker
that helps in the diagnosis

76. Inborn Error of Metabolism
 Disorders that have specific H-MRS patterns may
manifest as increase or absence of particular
metabolites.
 Specific biomarkers can be seen in phenylketonuria
(↑ phenylalanine7.36 ppm ), Canavan disease (↑
NAA), nonketotic hyperglycinemia (↑ glycine 3.55
ppm ), creatine deficiency (↓↓ Cr), and maple syrup
urine disease (branched-chain amino acids and keto
acids 0.9 ppm )