Brain Topography, Volume 5, Number 2, 1992

129

Nuclear Magnetic Resonance and the Brain E.R. Andrew

Summary: The first successful demonstrations of nudear magnetic resonance (NMR) in bulk matter were reported in 1946 (Bloch, Hansen and Packard 1946; Purcell, Torrey and Pound 1946). Since then NMR has become a widespread technique for investigating matter of all kinds. In the 1970's NMR was applied to living systems, including man, in 2 distinct approaches. One application was in the production of images (Lauterbur 1973), called Magnetic Resonance Imaging or MRI, and the other in the production of NMR spectra (Moon and Richards 1973; Hoult et al. 1974), called Magnetic Resonance Spectroscopy or MRS. By appropriate manipulation of the NMR signal an NMR image may be generated. This can be a 2D image of a single slice, or a set of 2D images of parallel slices, or a 3D image. 2D images may be obtained directly in any orientation, axial, coronal, sagittal. The method uses no ionizing radiation and is inherently safe. It is non-invasive, although paramagnetic solutions may be injected intravenously to improve contrast. MRI images observed in normal clinical practice are maps of the NMR signals from water and fat in the tissues; they depend on proton density, but also significantly on the relaxation times T1 and T2. Images can be provided of flow (MR angiography) and diffusion (free, restricted or anistropic). Images are typically 512x512 pixels with spatial resolution of about 0.5ram. The images can be correlated with anatomical structures and indeed MRI is a primary source of such structures with localization precision of 0.5ram as in CT. Normal imaging times are about 5mins, but fast images of lower resolution can be obtained in 50ms, enabling real time movie images to be generated. Recording sessions are typically 1hr. The NMR spectrum from living tissue gives a non-invasive measure of the concentration of each molecular species. Such spectra (MRS) provide information concerning the biochemistry of the body's metabolism and associated pathology. 31p spectra report concentrations of ATP, ADP, phosphocreatine, inorganic phosphate, other metabolites and also local pH. 1H spectra (with suppression of water and lipid responses) give spectra from lactate, NAA, choline, creatine and other components. Spectroscopic Imaging (SI) combines MRI and MRS to provide spectra simultaneously from an array of pixels or voxels, each usually several cm3 in size in an overall time of order 20 mins. This procedure provides a spatial map of the whole spectrum or individual maps of each molecular species. Two recent developments have demonstrated that NMR can provide functional mapping of the normal human brain, and map the response of the human cerebral cortex to physiological stimulation. Key words: NMR; MRI; MRS; Nuclear magnetic resonance; Brain metabolism; Imaging; Functional brain imaging; Spectroscopy.

Introduction This c o m m u n i c a t i o n describes the c o n t r i b u t i o n w h i c h N u c l e a r M a g n e t i c R e s o n a n c e (NMR) c a n m a k e to b r a i n i m a g i n g a n d b r a i n function. It will c o v e r b o t h m o d e s of o p e r a t i o n , n a m e l y M a g n e t i c R e s o n a n c e I m a g i n g (MRI) a n d M a g n e t i c R e s o n a n c e S p e c t r o s c o p y (MRS) a n d their c o m b i n a t i o n in M a g n e t i c R e s o n a n c e S p e c t r o s c o p y I m a g i n g (SI). N M R is c o n c e r n e d w i t h the nuclei of atoms, specifically w i t h a t o m i c n u c l e i w h i c h h a v e a n o n - z e r o m a g n e t i c m o m e n t ~t a n d s p i n I. T h e n u c l e u s o f a n o r d i n a r y h y d r o g e n a t o m 1H, a p r o t o n , p r o v i d e s a g o o d example. W h e n p l a c e d in a m a g n e t i c field B the p r o t o n precesses at a n a n g u l a r f r e q u e n c y c0=TB, w h e r e y is the n u c l e a r

Departments of Physics, Radiology and Nuclear Engineering Sciences, University of Florida, Gainesville, Florida, U.S.A. Accepted for publication: July 17, 1992. Support from NIH grant P41 RR02278 is gratefully acknowledged. Correspondence and reprint requests should be addressed to E.R. Andrew, Department of Physics, University of Florida, Gainesville, Florida, 32611, USA. Copyright © 1992 Human Sciences Press, Inc.

g y r o m a g n e t i c ratio. This result, d e r i v e d b y Sir J o s e p h L a r m o r in 1897, is a n essentially classical result. V i e w e d in e l e m e n t a r y q u a n t u m terms, the p r o t o n m a y be f o u n d in one of t w o eigenstates +~B. If w e a p p l y e l e c t r o m a g netic r a d i a t i o n of just s u c h a f r e q u e n c y v t h a t the q u a n t a exactly m a t c h the s e p a r a t i o n of the t w o states, t h e n a r e s o n a n t e x c h a n g e of e n e r g y c a n take p l a c e b e t w e e n the p r o t o n a n d the e l e c t r o m a g n e t i c field. H e n c e the n a m e Nuclear Magnetic Resonance. It r e q u i r e s

l e a d i n g to

m = 2~v = = yB

B since ~, = ~ t / ( h / 4 ~ )

We see that the NMR frequency from quantum m e c h a n i c s is exactly the s a m e as the classical L a r m o r f r e q u e n c y of precession. This h a p p y a g r e e m e n t enables us to l o o k at m a n y aspects of N M R classically. N o w c o n s i d e r a phial c o n t a i n i n g I m l w a t e r , s o m e 1023 p r o t o n s , in a m a g n e t i c field. I n e q u i l i b r i u m a small p r e p o n d e r a n c e of t h e m will line u p w i t h the field a n d give rise to a m a c r o s c o p i c m a g n e t i c m o m e n t , the n u c l e a r m a g n e t i z a t i o n M. If w e n o w a p p l y a n e l e c t r o m a g n e t i c

130

field at the Larmor frequency it turns M away from B, all the time precessing around it. Let us cut off the radiation after M has turned through 90 °. This is called a 90° pulse. If now we place a small coil of say six turns of copper wire around the phial of water, the precessing rotating magnetization M will generate, by Faraday's Law of Electromagnetic Induction, an electromotive force in the coil at the Larmor frequency. We can tune into this small signal with a radio receiver, amplify and detect it and display it on an oscillograph. This NMR signal can be obtained from all magnetic nuclei and every chemical element has at least one magnetic isotope. NMR was first successfully demonstrated in bulk matter by two independent groups simultaneously at Harvard (Purcell, Torrey and Pound 1946) and Stanford (Bloch, Hansen and Packard 1946). It had been sought unsuccessfully in Amsterdam by Gorter (1936) ten years earlier. To obtain NMR signals from the human brain in vivo the h u m a n subject is placed in a large magnet, usually of superconducting design with a I m bore, having a magnetic field typically in the range 0.5 to 2 T. For protons the corresponding NMR frequency range is 21.3 to 85.2 MHz.

Nuclear Magnetic Resonance Imaging (MRI) We now consider how NMR can be used to obtain an image of a structured heterogeneous object such as the human head. An outline is given here and further details of both MRI and MRS and their applications may be found in the book by Andrew, Bydder et al. (1990). First we define a slice in the head and then image the distribution of protons in that slice. To define the slice we apply a magnetic field gradient along the Z direction and irradiate the head with a 90 ° resonant NMR pulse of narrow spectral width, which therefore excites only those °protons in a thin slice which are in resonance. This is selective excitation (Garroway, Grannell and Mansfield 1974). We now switch the gradient quickly into the plane of the slice along X, and allow the proton NMR signal to evolve in the presence of this gradient. Each volume element in the slice is spatially encoded and its NMR frequency depends on its position, its X coordinate. After time tx we switch the gradient along the Y direction. Each volume element now has an NMR frequency which depends on its Y coordinate. The NMR signal is now recorded during the time ty that the Y gradient is applied. It is recorded to say 256 data points. The process is repeated with an increased value of tx; indeed it is repeated 256 times yielding a matrix of 256x256 data values. The instrument must have its own dedicated computer which instructs the system, collects and stores

Andrew

the data, manipulates the data and carries out subsequent calculations. The computer now subjects this array of data to a 2D Fourier transform and constructs a 2D image. This procedure is called the 2D Fourier Imaging Method, due to Kumar, Welti and Ernst (1975). The very first NMR images were obtained by Lauterbur (1973), who introduced the essential step of a spatially-encoding field gradient, but generated his images by a somewhat different procedure. A variant of the 2D Fourier imaging method produces a simultaneous set of images of parallel slices; this is the multislice Fourier Imaging method. Alternatively one may dispense with slice selection and apply the gradient successively along X,Y,Z for times tx,ty,tz and then apply a 3D Fourier transform; this generates a 3D NMR image from which the computer may select and display any desired 2D image. This is the 3D Fourier Imaging Method. A feature of MRI (in contrast with CT X-ray) is that sagittal and coronal images may be obtained directly and with equal ease to transverse images. We should note that MRI does not use ionizing radiation and if used properly it is a non-hazardous modality of internal structural mapping. In its basic operation it is non-invasive. However it is often possible to improve the contrast in an image by intravenous injection of a paramagnetic solution such as Gd DTPA. The large DTPA molecule does not readily pass through the bloodbrain barrier but can more readily enter tumors and other lesions altering the local proton NMR signal and rendering the lesion more visible. Although the best spatial resolution in an image of the whole head is about 0.5 mm, better resolution of a small region of the head may be obtained, for example 0.1 m m in the eye. Moreover it is possible to image quite small structured objects of order 1 m m in size, for example simple cellular systems, and obtain a spatial resolution down to 4~t ; such developments are referred to as NMR microimaging. MRI images observed in normal clinical practice are maps of the proton NMR signals from water and fat in the tissues. Images can also be provided of flow (MR a n g i o g r a p h y ) a n d d i f f u s i o n (free, r e s t r i c t e d or anisotropic). Images are typically 512x512 picture elements (pixels) with spatial resolution of about 0.5 mm. The images can be correlated with anatomical structures, and indeed MRI is a primary source of such structures with localization precision of at least 1 mm. Normal imaging times for a multislice set are about 5 min, but faster images of lower resolution can be obtained in 5 ms enabling real time movie images to be generated. Recording sessions are typically 1 hr. MRI is excellent for the morphology of the head, the flow of blood and CSF, cerebral blood volume (CBV) and diffusion, for the clinical study of tumors, ischemia, hemorrhage, muscular

NMR and the Brain

sclerosis and other pathology of the head. The great majority of MRI has been done with 1H nuclei, but some work has been done with other nuclei such as 13C, 31p and 23Na.

Nuclear Magnetic Resonance Spectroscopy (MRS) On close examination it is found that the NMR response from a liquid is not at one well-defined frequency, but comes from a close cluster of discrete frequencies. This fine structure has its origin in the molecular electrons surrounding the resonant nuclei, and endows each molecule with its own recognizable NMR response. The NMR spectrum from living tissue thus gives a noninvasive measure of the concentration of each molecular species. Such spectra (MRS) provide information concerning the biochemistry of the body's metabolism and associated pathology. The first in vivo NMR spectra were recorded by Moon and Richards (1973) and by Hoult et al. (1974). The three nuclei most commonly used in MRS are 1H, 13C and 31p. The two strong proton NMR responses utilized in MRI, namely from water and fat, are of little use in MRS, but useful spectral information comes from the much weaker responses from lactate, N-acetyl-aspartate, choline, creatine and other components. 31p spectra give clear responses from ATP, ADP, phosphocreatine, inorganic phosphate and other metabolites and also local pH values. It is important to know from which anatomic region the NMR spectra are being generated and many methods have been developed to answer this problem with fair success. The metabolites whose NMR spectra are recorded have a weak concentration in tissues, typically 10-4 of that of water. Consequently the smallest volume of tissue which will generate acceptable 1H and 31p spectra is typically 10 ml. Acquisition times for MRS tend to be somewhat longer than for MRI, typically 15 min and recording sessions last an hr or more. The natural abundances of 1H and 31p are 99.8% and 100%. By contrast that of 13C is only 1.1% and consequently the NMR signals from tissues are unacceptably weak. It is therefore often desirable to introduce metabolites in which 13C has been enriched, for example in glucose. Because of the lower sensitivity and consequently poorer spatial resolution of MRS, it has not followed MRI into hospitals as a general modality of radiology. It has clinical applications but at present fs primarily an investigative research technique giving valuable information concerning normal and abnormal biochemistry, physiology and pathology of the human body. Spectroscopic volumes can be correlated with anatomy using MRI.

131

NMR Spectroscopic Imaging The procedures of MRI and MRS can be combined. One first selects a slice, applies a magnetic field gradient successively along X and Y as in conventional MRI and then goes on to collect the NMR signal without gradient. The periods tx and ty during which the X and Y gradients are applied are both independently incremented n times and a 3D Fourier transform is applied to the accumulated data. This procedure provides a spatial map of the whole spectrum, or individual maps of each molecular species. The pixels tend to be rather large, of order 1-3 cm. An MRI image of the same slice taken immediately prior to the SI series enables all the SI pixels to be located anatomically.

Nature of the NMR Signals From the Brain Recorded It will be seen from the discussion of MRI and MRS that the NMR signals of from the brain which generate the images and spectra have no direct connection with brain function or activity. They are of the same kind as the NMR signals recorded from inanimate matter. Nevertheless the function and dysfunction of the brain can certainly affect the NMR response in a measurable way. Tumors and other lesions alter brain morphology in a clearly recognizable manner making MRI a valuable tool in clinical diagnosis. Epileptic activity raises the lactic acid concentration in the brain as evidenced by an increase in the intensity of the lactic acid doublet in the 1H MRS spectrum. Matthews and Arnold (1989) found a tenfold increase in lactate near the left sylvian fissure compared with the right sylvian fissure in a ten year old girl with focal epilepsy; other metabolites were unaffected. Several workers have noted that the onset of stroke increases lactic acid concentration in the affected area. An example is van Rijen et al. (1989a) who found a lactate increase up to four times in the abnormal area determined by MRI. In the clinical recovery phase up to 48 days lactate remained elevated suggesting ongoing glycolysis in the affected region. Electric shock applied to the cortex of rabbits raised the lactate concentration (Petroff et al. 1989). Hyperventilation of normal healthy volunteers with 2s breathing period was found to double the lactate concentration (van Rijen et al. 1989b). After cessation of rapid breathing the lactate concentration returned to normal in about 15 rain. Of the examples just given, epileptic seizures and strokes represent severe pathological provocations to the brain. An electric shock to the cortex is a severe physiological provocation. We may ask what is the effect of much milder physiological stimuli such as those of our ordinary senses?

132

MRS Determination of Lactate Rise by Optical Stimulation Prichard et al. (1991) used proton MRS to measure increased lactate concentration in the human visual cortex when stimulated b y a flashing light. The lactate concentration was monitored through observation of the lactate methyl doublet. Proton spectra were collected from a 13 c m ° v o l u m e of the visual cortex before, during and after photic stimulation. For five human subjects, stimulated for 12-18 minutes, the average increase of lactate concentration was 54%, rising from 0.7 mM to 1.1 mM. The lactate concentration declined after initial stimulation and returned to normal in about 40 minutes. It is clear that the lactate response can be mapped throughout the cortex either by successive measurement of defined volumes or simultaneously by the spectroscopic imaging method. These results have been confirmed by one other group (Sappey-Marinier et al. 1990) though a third group (Merboldt et al. 1991, 1992) found no consistent increase in lactate concentration but did find a 50% reduction in glucose concentration. Similar MRS cortical mapping can be expected before long for auditory, motor and other stimuli.

MRI Functional Mapping of the Human Visual Cortex MRI also can measure the response to optical stimulation by a flashing light and does so faster and with greater spatial resolution than MRS. The pioneer work was carried out by Belliveau et al. (1991). The principle here is that intravenous injection of a paramagnetic contrast agent such as Gd DTPA causes the paramagnetic Gd ions to pass t h r o u g h all b l o o d vessels large and small throughout the head and body. The magnetic field gradients generated b y these magnetic ions dephase the precessing proton magnetic moments in their vicinity and reduce the N M R signals from surrounding tissues. The integrated reduction of NMR signal is a measure of local blood volume. Using a fast MRI technique images of the visual cortex were obtained every 750 ms for 45 sec after injection of the contrast agent. The integrated NMR signal from each voxel was then displayed as an image. The procedure was carried out twice (a) with the human subject resting and (b) with the light flashing. The difference image obtained by subtraction of (a) from (b) gave an image of increased local CBV caused by the photic stimulation. The image slice thickness was either 8 or 10 m m and the in-plane voxel size either 1.5xl.5 mm or 3x3 mrn, giving voxels ranging from 18 to 90 mm 3. The average increase in CBV for seven human subjects was 32 + 10% in the affected region. These two papers using MRS (Prichard et al. 1991) and

Andrew

MRI (Belliveau et al. 1991) have opened a new chapter in the use of NMR and a new contribution of NMR to brain function. Because MRI is measuring the much stronger NMR signals from water it is much faster with much better spatial resolution. On the other hand MRS, responding to the weak concentration of lactic acid, is metabolite specific. Both procedures provide a direct map of stimulated activity in the human cortex without use of ionizing radiation.

References Andrew, E.R., Bydder, G., Griffiths, J., Iles, R., Styles, P. Clinical Magnetic Resonance: Imaging and Spectroscopy. John Wiley and Sons, Chichester and New York, 1990. Belliveau, J.W., Kennedy, D.N., McKinstry, R.C., Buchbinder, B.R., Weisskoff, R.M., Cohen, M.S., Vevea, J.M., Brady, T.J., Rosen, B.R. Functional mapping of the human visual cortex by magnetic resonance imaging. Science 1991, 254: 716-719. Bloch, F., Hansen, W.W., Packard, M.E. Nuclear induction. Phys Rev 1946, 69: 127. Garroway, A.N., Grannell, P.K., Mansfield, P. Image formation in NMR by a selective irradiative process. J Phys C: Solid State Phys 1974, 7:L457-462 Gorter, C.J. Negative result of an attempt to detect nuclear magnetic spins. Physica, 1936, 3: 995-998. Hoult, D.I., Busby, S.J.W., Gadian, D.G., Radda, G.K., Richards, R.E., Seeley, P.J. Observation of tissue metabolites using 31P nuclear magnetic resonance. Nature 1974, 252: 285-287. Kumar, A., Welti, D., Ernst, R.R. NMR Fourier Zeugmatography. J Mag Res 1975, 18: 69-83. Lauterbur, P.C. Image formation by induced local interactions: Examples employing nuclear magnetic resonance. Nature 1973, 252: 190-191. Matthews, P.M., Arnold, D.L. In vivo proton magnetic resonance spectroscopy in the study of focal epilepsy in man. SMRM Abstracts 371, 1989. Merboldt, K.D., Bruhn, H., Gyngell, M.L., Hanicke, W., Michaelis, T., Frahm, J. Variability of lactate in normal human brain in vivo. Localized proton MRS during rest and photic stimulation SMRM Abstracts 392, 1991. Merboldt, K.D., Bruhn, H., Hanicke, W., Michaelis, T., Frahm, J. Decrease of glucose in the human visual cortex during photic stimulation. Mag Res Med 1992, 25: 187-194. Moon, R.B. and Richards, J.H. Determination of intracellular pH by 31p magnetic resonance. J Biol Chem 1973, 248: 72767278. Petroff, O.A.C., Novotny, E.J., Avison, M.J., Rothman, D.L., Shulman, R.G., Prichard, J.W. Cerebral lactate turnover after electroshock by proton observed carbon decoupled spectroscopy. SMRM Abstracts 332, 1989. Prichard, J., Rothman, D., Novotny, E., Petroff, O., Kuwabara, T., Avison, M., Howseman, A., Hanstock, C., Shulman, R. Lactate rise detected by 1H NMR in human visual cortex during physiologic stimulation. Proc Natl Acad Sci USA 1991, 88: 5829-5831. Purcell, E.M., Torrey, H.C,. Pound, R.V. Resonance absorption by nuclear magnetic moments in a solid. Phys Rev 1946, 69: 37-38.

NMR and the Brain

van Rijen, P.C., Luyten, P.R., den Hollander, J.A., Tulleken, C.A.F. Prolonged elevation of cerebral lactate detected with 1H NMR spectroscopy in patients with focal cerebral ischemia. SMRM Abstracts 374, 1989a. van Rijen, P.C., Luyten, P.R., van der Sprenkel, J.W.B., Kraaier, V., Van Huffelen, A.C., Tulleken, C.A.F., den Hollander, J.A.

133

1H and 31p NMR measurement of cerebral lactate, high-energy phosphate levels, and pH in humans during voluntary hyperventialation. Mag Res Med 1989b, 10: 182-193. Sappey-Marinier, D., Calabrese, G., Hugg, J., Deicken, R., Fein, G., Weiner, M. Increased lactate in human visual cortex during photic stimulation. SMRM Abstracts 106, 1990.

Nuclear magnetic resonance and the brain.

The first successful demonstrations of nuclear magnetic resonance (NMR) in bulk matter were reported in 1946 (Bloch, Hansen and Packard 1946; Purcell,...
503KB Sizes 0 Downloads 0 Views