Electroencephalography and clinical Neurophysiology , 85 (1992) 215-219 © 1992 Elsevier Scientific Publishers Ireland, Ltd. 0924-980X/92/$05.00
215
ELMOCO 91121
M a g n e t i c brain s t i m u l a t i o n a n d brain size: relevance to a n i m a l studies J.D. Weissman b, C.M. Epstein a,b and K.R. Davey a,c a Rehabilitation Research and Development Center, Atlanta VA Medical Center, Decatur, GA (U.S.A.), b Department of Neurology, Emory University School of Medicine, Atlanta, GA (U.S.A.), and c Department of Electrical Engineering, Georgia Institute of Technology, Atlanta, GA (U.S.A.) (Accepted for publication: 30 January 1992)
Summary
Magnetic brain stimulation (MBS) is widely used for the investigation of brain function in man, but there have been only a few reports of its safety in animals. These results were predominantly benign, but the effectiveness of stimulation in animals is unclear. Because the stimulators produced obvious motor effects in humans, or had a comparable peak magnetic field strength, they were assumed to produce comparable electric field intensities and neuronal effects in animal brains. We tested this assumption using 3 stimulus coils of different sizes and design, plus 6 saline-filled spheres that spanned a range of volume from 0.5 to 1800 ml. The induced electric field diminished monotonically with decreasing radius, by factors of 4.7-6.2 at the extremes of size. Comparable results were found using a mathematical model. These results suggest that the efficiency of magnetic stimulation is drastically reduced in smaller brains, and that threshold and safety studies in some animal models may not be valid.
Key words: Magnetic brain stimulation; Animal studies; Electric field; Magnetic field; Safety
Magnetic brain stimulation (MBS) is increasingly used for the investigation of motor pathways in patients and normal subjects (Barker et al. 1986; Schriefer et al. 1987, 1989; Claus et al. 1988; Hallett and Cohen 1989; Miiller et al. 1991). With a limited number of single stimuli, its application is generally considered safe (Barker et al. 1990; Tassinari et al. 1990). Brief trials of MBS in humans have not been associated with EEG (Levy et al. 1990), SPECT (Dressier et al. 1990), neuropsychological (Bridgers and Delaney 1989) or hormonal changes (Thomas et al. 1991). Possible induction of seizures has been reported rarely (H6mberg and Netz 1989). Trains of magnetic stimuli may provoke typical seizures in patients with pre-existing epilepsy (Hufnagel et al. 1990), and early reports of rapid repetitive MBS suggest a greater seizure-evoking ability (Pascual-Leone et al. 1991). In general, however, human safety studies have used limited numbers of subjects and stimuli, and follow-up has been brief. Pathologic examination has been possible only in patients about to undergo brain resection for the treatment of intractable epilepsy (Pascual-Leone et al. 1991). Surprisingly few investigators have described long-term effects of MBS in experimental animals. Small animals can be exposed to more stimuli
Correspondence to: Charles M. Epstein, M.D., Rehabilitation Research and Development Center, Atlanta VA Medical Center, 1670 Clairmont Road, Decatur, GA 30033 (U.S.A.).
than humans, at all stages of development and for a greater fraction of their life span. Histological analysis might detect subtle long-term effects that are difficult to recognize in human patients. But only a few, mostly preliminary, reports of animal safety studies are available for MBS. Although vacuolar changes are described by Matsumiya et al. (1992), others have reported no changes (Mano et al. 1989; Hersh et al. 1990; Kling et al. 1990; Sgro et al. 1990). Kling et al. (1990) felt that the power of their magnetic stimulator would be greater in rats because of the thinner skull, but currently available stimulators may be less effective in animals than in man. The practice of using the peak magnetic field strength as the sole measure of magnetic stimulating power may be misleading. Although responses similar to those of humans have been reported in monkeys (Amassian et al. 1987), dogs (Heckmann et al. 1989) and cats (Fujiki et al. 1990), they have been absent in mice (Hersh et al. 1990) and guinea pigs (R. Gilmore, personal communication). In our experience and elsewhere (from conversations with other investigators) it seems difficult to determine if motor responses to MBS are present in small animals or, if present, are due to activation at a central rather than a peripheral site. Even the body jerking described by Kling et al. (1990) in rats might be due to activation of lower-threshold peripheral nerve or skeletal muscle rather than brain. The apparent differences among species may be due simply to variations in the organization of the motor system. An
216 alternative possibility, however, is that magnetic stimulators are less efficient at inducing current in the confines of a smaller head. We report here studies which indicate that magnetic stimulus coils appropriate for humans produce substantially weaker effects in the much smaller brains of common experimental animals.
Methods
Magnetic stimulator and coils The stimulator and two of the coils, which had a modified "butterfly" configuration, have previously been described (Epstein et al. 1990). The larger and smaller butterfly coils had central segments measuring 10 and 5.3 cm respectively. For these studies we added a circular coil of comparable inductance (10/zhenrys), with an inside diameter of 8.5 cm and outside diameter of 11 cm. Brain models and electric field measurements Theoretical analyses have shown that, for modeling the currents induced during magnetic brain stimulation, the brain and skull can be represented as a homogenous conductor surrounded by an insulator, with boundaries corresponding to the inner table of the skull (Cohen and Cuffin 1991). H~im~il~iinen and Savas (1989) used a similar model for magnetoencephalography, which may be applied to magnetic brain stimulation by invoking the reciprocity theorem. We therefore constructed 6 thin-walled spherical models with internal radii of 0.5-7.5 cm. Corresponding volumes were 0.5-1800 ml, spanning the brain size range from mouse to human. To normalize wall thickness and represent the skull, extra layers of hard rubber insulation were added between some of the spherical models and the stimulus coil. The gap from stimulator to saline boundary was 5 mm for the 4 largest spheres, and 2.5 mm for the two smallest. We tested 3 different conductive media: 0.9% saline (Maccabee et al. 1990), 0.08% normal saline (Geddes and Bourland 1990), and a saturated NaC1 solution approximately equal to 4 N. The spheres were filled entirely with saline solution and a test probe was inserted through a small hole in the top, with the stimulus coil directly beneath the sphere. Measurements were made with a variety of twisted pair (Geddes and Bourland 1990) and concentric (Maccabee et al. 1990) probes, using both bare copper and silver-silver chloride electrodes. Probes were placed tangent to the inner wall of the sphere 2 mm inside the saline boundary. For the butterfly coils, all measurements were made above the middle of the central segment, where the induced electric field was known to be largest (R6sler et al. 1989; L.G. Cohen et al. 1990; Epstein et al. 1990; Maccabee et al. 1990). For the round coil, measurements were made 2 mm from
J.D. WEISSMAN ET AL. the inner wall of the sphere at the point where it contacted the coil. To avoid distortion by current within the probe, the input to a storage oscilloscope was fed through a high impedance unity-gain field-effect transistor (FET) amplifier. Capacitive effects were minimized by a 20 × 20 cm copper screen Faraday shield between the coil and the model head. Throughout the course of the measurements we regularly reversed the direction of the probe to test for polarization effects and turned the probe at right angles to the coil windings to check for effects of capacitance. In the absence of significant eddy currents, the induced electric field should be proportional to the rate of change of the magnetic flux captured by the target, which is in turn a function of its size (Tenforde 1986). To test this hypothesis we constructed a series of round single-turn search loops, with radii matching the positions of the probes within the saline-filled spheres.
Mathematical modeling of induced fields The effect of target size was calculated theoretically with a model of homogenous spheres centered over a round stimulus coil, using elliptical integrals to predict the field produced by the coil and the Gauss-Legendre integration technique (Abramowitz and Stegun 1970; Davey et al. 1991) to determine the flux at the test region. Integrating this field yields the flux. Our model coil had a radius of 5 cm with negligible thickness and carried 1000 A turns with a rise time of 100/xsec. The electric fields were arrived at by integrating the magnetic flux over an area within the volume of interest (at a depth equal to 10% of the radius) using the integral form of Maxwell's equations.
Results
Electric field measurements Preliminary studies showed that the electric field strength at equivalent positions was identical for 0.08% saline, normal saline, and the saturated saline solution. This was the expected result in the absence of significant eddy currents (Tenforde 1986). A high noise level was obtained initially from the 3 smallest spheres and dilute saline. Therefore saturated saline was used with the smaller spheres. Comparable electric field measurements were obtained with copper and silver-silver chloride electrodes of both varieties; polarization effects were absent with the brief pulses and very high impedance amplifier used. However, the bare twistedwire pair was easiest to position in the smaller spheres. Consequently a single twisted-wire pair with a gap of 4 mm was used in one consecutive session to obtain all the measurements of Fig. 1. Reversal of probe direction invariably showed that amplitude was unchanged. Wave shape was identical to that measured with a
MAGNETIC STIMULATION AND BRAIN SIZE
217
search coil in air. With the probe oriented at right angles to the stimulator coil windings, the measured potential was always less than 25% of that recorded in the parallel measurements. This and the fact that the induced voltage followed the rate of change of the coil current excluded substantial capacitive effects. Fig. 1 shows the relationship between the diameters of the 6 test spheres and the potentials measured 2 mm inside the inner wall. For all 3 coils, the potentials fall substantially with the radius of the sphere, by factors of 4.7-6.2 at the extremes of size. These ratios are probably conservative because of residual capacitance, with y-intercepts of approximately 0.1 V / c m . For comparison, brain weights of human (1400 g, Neal and Rand 1936), orangutan (350 g, Neal and Rand 1936) and rabbit (30 g, Crosby and Schnitzlein 1982) were taken from standard sources. The weights for rat and mouse brain were measured in our lab and were 2 and 0.5 g, respectively. Assuming a specific gravity near 1 and spherical shape, equivalent radii were calculated as shown in Fig. 1. Note that for the smallest targets the electric field is about twice as strong with the smaller butterfly coil, which is half the size of the larger one. We also compared measurements made with the sphere centered over the edge of the round coil, tangent to its plane, with measurements taken at the center of the coil. These studies were performed with a 0.25 cm concentric probe. For the 2 smallest spheres, roughly similar in size to the brains of a mouse and a rat, the measured potentials were 30% higher with the target in the center of the coil than at the edge. However, they were still several times lower than for the largest sphere.
__A-- L a r g e
Small Butterfly
.o.
Round
Butterfly
1.20 E
1.00
Spherical
Radius,
2.00
1.50
°E
>
/
1.00
/
0.50
/
f
0.00 0
1
2
3
4
5
6
test region radius cm
Fig. 2. Mathematical prediction of peak electric field at or above the center of a circular stimulus coil, calculated for spherical targets of 0-6 cm radius.
The peak voltages induced in the single-turn circular search coils were highly correlated with the radii of the respective head models, with correlation coefficients r--0.98, 0.95 and 0.99 for the small butterfly, large butterfly and round coil, respectively (all P < 0.01).
Mathematical modeling We calculated the induced fields for the case of targets smaller than the stimulus coil along a line passing through the center of the coil perpendicular to its plane. For radii greater than 5 cm, the target could no longer lie in the center of the coil, but contacted it at all points. Fig. 2 illustrates the calculated results. As expected, the induced electric field approaches zero for a test region of zero radius. The calculated field in the vicinity of 5 cm radius is probably distorted, due to the implicit assumptions of negligible coil thickness and negligible skull thickness. Otherwise the mathematical model shows the same behavior as the empirical measurements in Fig. 1.
cm
0
j 0.80
C
....
~
Discussion
.... ; ~
0.60
_u 0.40 LLI
/
0,20 0.00
'
0
1
::
i r~lt mouse rabbit
'
'
'
~
~
J
J
2
3
4
5
6
7
8
orangutan
human
Fig. 1. Peak electric fields, measured with a twisted-wire probe, using 3 different magnetic stimulus coils and 6 spherical head models of different sizes. Representative brain volumes of 5 species are shown for comparison according to their equivalent radius.
These results indicate that the efficiency of MBS drops substantially when the target becomes smaller than the stimulus coil. Experimentally this relationship holds true both for circular coils, which are the most commonly used, and for butterfly coils. The electric field difference which drives current flow is dependent on the changing magnetic flux through the target. A smaller target will capture a proportionately smaller fraction of the total flux. Even if, within the brain, membrane depolarization depends on the electric field gradient (Reilly 1989; Roth and Basser 1990), the gradient decreases as the target volume decreases. This
218
can be analytically shown with the target centered on the coil axis and implicitly shown with the target offaxis. As the volume of the brain falls from that of a human to that of a mouse, the magnitude of the induced field drops by a ratio of at least 5 to 1. But with magnetic stimulation in humans, the threshold of movement typically occurs around 50% of full output (Amassian and Cracco 1987; Amassian et al. 1988; Ackermann et al. 1991; Miiller et al. 1991). If the neuronal thresholds in different species are comparable, even full output would seem insufficient for effective stimulation in small rodents. Our measurements support the technique chosen by Sgro et al. (1990), who stimulated the rat brain inside the center of a round stimulus coil. This is not possible for the human head and body, whose dimensions are larger than the coil itself. Nonetheless, stimulus efficiency is still lower than for a larger volume. Maccabee et al. (1991) reported analogous results with stimulation of sensory fibers in the fingers. They found that small volume conductors such as the fingers required targeting near the center of a round coil instead of under the windings in order to elicit a sensory action potential. Figs. 1 and 2 suggest a simple rule-of-thumb for magnetic brain stimulation with heads of different sizes: the stimulus coil should be comparable in size to the brain, or smaller for greater focality. But for small animals this principle collides immediately with a fundamental physical difficulty. Smaller coils are subject to progressively greater thermal and mechanical stresses and may fail catastrophically (Fortescue and Bickford 1990; Cohen and Cuffin 1991). Practical coils for small brain stimulation may require different geometries or iron cores and may be a challenge to design. The effects of brain size also may account in part for the decrease in magnetic stimulation threshold during human maturation (Koh and Eyre 1988). Similarly, any comparison of the thresholds for magnetic stimulation in brains of different species must take into account the changing efficacy of stimulation associated with differences in brain size. Meanwhile animal safety studies of MBS remain desirable but limited, and these considerations may help to explain the paucity of reports. Our results raise doubts about the validity of some studies conducted thus far, especially those in which large stimulating coils designed for human studies are applied to small animal brains or tissue cultures in small containers. References Abramowitz, M. and Stegun, C. Handbook of Mathematical Functions, 9th Edition. Dover Publications, 1970: 591-592. Ackermann, H., Scholz, E., Koehler, W. and Dichgans, J. Influence
J.D. WE1SSMAN ET AL. of posture and voluntary background contraction upon compound muscle action potentials from anterior tibial and soleus muscle following transcranial magnetic stimulation. Electroenceph. clin. Neurophysiol., 1991, 81: 71-80. Amassian, V.E. and Cracco, R.Q. Human cerebral cortical responses to contralateral transcranial stimulation. Neurosurgery, 1987, 20: 148-155. Amassian, V.E., Quirk, G.J. and Stewart, M. Magnetic coil versus electrical stimulation of monkey motor cortex. J. Physiol. (Lond.), 1987, 304: lt9P. Amassian, V.E., Bigland-Ritchie, B., Cracco, R.Q. and Maccabee, P.J. Motor unit fields mapped by focal magnetic coil stimulation of human motor cortex. J. Physiol. (Lond.), 1988, 420: 20P. Barker, A.T., Freeston, I.L., Jalinous, R. and Jarratt, J.A. Clinical evaluation of conduction time measurements in central motor pathways using magnetic stimulation of the human brain. Lancet, 1986, i: 1325-1326. Barker, A.T., Freeston, I.L., Jarratt, J.A. and Jalinous, R. Magnetic stimulation of the human nervous system: an introduction and basic principles. In: S. Chokroverty (Ed.), Magnetic Stimulation in Clinical Neurophysiology. Butterworth, Boston, MA, 1990: 55-72. Bridgers, S.L. and Delaney, R.C. Transcranial magnetic stimulation: an assessment of cognitive and other cerebral effects. Neurology, 1989, 39: 417-419. Cadwell, J. Principles of magnetoelectric stimulation. In: S. Chokroverty (Ed.), Magnetic Stimulation in Clinical Neurophysiology. Butterworth, Boston, MA, 1990: 13-32. Claus, D., Harding, A.E., Hess, C.W., Mills, K.R., Murray, N.M. and Thomas, P.K. Central motor conduction in degenerative ataxic disorders: a magnetic stimulation study. J. Neurol. Neurosurg. Psychiat., 1988, 51: 790-795. Cohen, D. and Cuffin, B.N. Developing a more focal magnetic stimulator. Part 1. Some basic principles. J. Clin. Neurophysiol., 1991, 8: 102-111. Cohen, L.G., Roth, B.J., Nilsson, J., Dang, N., Panizza, M., Bandinelli, S., Friauf, W. and Hallett, M. Effects of coil design on delivery of focal magnetic stimulation. Technical considerations. Electroenceph. clin. Neurophysiol., 1990, 75: 350-357. Crosby, E.C. and Schnitzlein, H.N. Correlative Neuroanatomy o1 the Vertebrate Telencephalon. McMillan Publishing Co., New York, 1982:414 pp. Davey, K.R., Chin, H.C. and Epstein, C.M. Prediction of magnetically induced electric fields in biological tissue. IEEE Trans. Biomed. Eng., 1991, 38: 418-426. Dressier, D., Voth, E., Feldmann, M. and Benecke, R. Safety aspects of transcranial brain stimulation in man tested by single photon emission computed tomography. Neurosci. Lett., 1990, 119: 153155. Epstein, C.M., Schwartzenberg, D.G., Davey, K.R. and Sudderth, D.B. Localizing the site of magnetic stimulation in humans. Neurology, 1990, 40: 666-670. Fortescue, P. and Bickford, R. The design of electromagnetic stimulators. In: S. Chokroverty (Ed.), Magnetic Stimulation in Clinical Neurophysiology. Butterworth, Boston, MA, 1990: 45-53. Fujiki, M., Isono, M. and Hori, S. Spinal and muscle motor evoked potentials following magnetic stimulation in cats. Neurol. Med. Chir. (Tokyo), 1990, 30: 234-241. Geddes, L.A. and Bourland, J.D. Fundamentals of eddy-current (magnetic) stimulation. In: S. Chokroverty (Ed.), Magnetic Stimulation in Clinical Neurophysiology. Butterworth, Boston, MA, 1990: 33-43. Hallett, M. and Cohen, L.G. Magnetism, a new method for stimulation of nerve and brain. JAMA, 1989, 262: 538-541. H~im~il~iinen, M.S. and Savas, J. Realistic conductivity geometry model of the human head for interpretation of neuromagr~etic data. IEEE Trans. Biomed. Eng., 1989, 36: 165-171.
MAGNETIC STIMULATION AND BRAIN SIZE Heckmann, R., Hess, C.W., Hogg, H.P., Ludin, H.P. and Wiestner, T. Transcranial magnetic stimulation of the motor cortex and percutaneous magnetic stimulation of the peripheral nervous structures in the dog. Schweiz. Arch. Tierheilk., 1989, 131: 341350. Hersh, S.M., Green, R.C., Weissman, J.D., Rees, H.C., Davey, K.R., Epstein, C.M. and Bakay, R.A.E. Biological consequences of transcranial magnetic stimulation in the mouse. Abstract presented at Society for Neuroscience 1990 Annual Meeting, St. Louis, MO, November 1, 1990. H6mberg, V. and Netz, J. Generalized seizures induced by transcranial magnetic stimulation of motor cortex. Lancet, 1989, ii: 1223. Hufnagel, A., Elger, C.E., Durwen, H.F., Boker, D.K. and Entzian, W. Activation of the epileptic focus by transcranial magnetic stimulation of the human brain. Ann. Neurol., 1990, 27: 49-60. K.ling, J.W., Yarita, M., Yamamoto, T. and Matsumiya, Y. Memory for conditioned taste aversions is diminished by transcranial magnetic stimulation. Physiol. Behav., 1990, 48: 713-717. Koh, T.H. and Eyre, J.A. Maturation of corticospinal tracts assessed by electromagnetic stimulation of the motor cortex. Arch. Dis. Childh., 1988, 63: 1347-1352. Levy, W.J., Oro, J., Tucker, D. and Haghighi, S. Safety studies of electrical and magnetic stimulation for the production of motor evoked potentials. In: S. Chokroverty (Ed.), Magnetic Stimulation in Clinical Neurophysiology. Butterworth, Boston, MA, 1990: 165-172. Maccabee, P.J., Eberle, L.P., Amassian, V.E., Cracco, R.Q., Rudell, A. and Jayachandra, M. Spatial distribution of the electric field induced in volume by round and figure 8 magnetic coils: relevance to activation of sensory nerve fibers. Electroenceph. clin. Neurophysiol., 1990, 76: 131-141. Maccabee, P.J., Amassian, V.E., Cracco, R.Q., Eberle, L.P. and Rudell, A.P. Mechanisms of peripheral nervous system stimulation using the magnetic coil. In: W.J. Levy, R.Q. Cracco, A.T. Barker and J.C. RothweU (Eds.), Magnetic Motor Stimulation. Electroenceph. clin. Neurophysiol., Suppl. 43. Elsevier, Amsterdam, 1991: 340-357. Mano, Y., Funakawa, I., Nakamuro, T., Takayanagi, T. and Matsui, K. The kinesiological, chemical and pathological analysis in pulsed magnetic stimulation to the brain. Rinsho Shinkeigaku, 1989, 29: 982-988. Matsumiya, Y., Yamamota, T., Yarita, M., Miyauchi, S. and Kling, J.W. Cortical damage by magnetic stimulation. J. Clin. Neurophysiol., 1992, in press.
219 Miiller, K., H6mberg, V. and Lenard, H.-G. Magnetic stimulation of motor cortex and nerve roots in children. Maturation of corticomotoneuronal projections. Electroenceph. clin. Neurophysiol., 1991, 81: 63-70. Neal, H.V. and Rand, H.W. Comparative Anatomy. Blaki~ton Publishing Co., Philadelphia, PA, 1936:505 pp. Pascual-Leone, A., Gates, J.R. and Dhuna, A. Induction of speech arrest and counting errors with rapid transcranial magnetic stimulation. Neurology, 1991, 41: 697-702. Reilly, J.P. Peripheral nerve stimulation by induced electric currents: exposure to time-varying magnetic fields. Med. Biol. Eng. Comput., 1989, 27: 101-110. Roth, B.J. and Basser, P.J. A model of the stimulation of nerve fiber by electromagnetic induction. IEEE Trans. Biomed. Eng., 1990, 37: 588-597. Roth, B.J., Saypol, J.M., Hallett, M. and Cohen, L.G. A theoretical calculation of the electrical field induced in the cortex during magnetic stimulation. Electroenceph. clin. Neurophysiol., 1991, 81: 47-56. R6sler, K.M., Hess, C.W., Heckmann, R. and Ludin, H.P. Significance of shape and size of the stimulating coil in magnetic stimulation of the human motor cortex. Neurosci. Lett., 1989, 100: 347-352. Schriefer, T.N., Mills, K.R., Murray, N.M.F. and Hess, C.W. Magnetic brain stimulation in functional weakness. Muscle Nerve, 1987, 10:463 (abstract). Schriefer, T.N., Hess, C.W., Mills, K.R. and Murray, N.M. Central motor conduction studies in motor neurone disease using magnetic brain stimulation. Electroenceph. clin. Neurophysiol., 1989, 74: 431-437. Sgro, J.A., Stanton, P.C., Emerson, R.G. and Ghatak, N.R. Repetitive high magnetic field stimulation: the effect on rat brain. Electroenceph. clin. Neurophysiol., 1991, 78: 24P. Tassinari, C.A., Michelucci, R., Forti, A., Plasmati, R., Troni, W., Salvi, F., Blanco, M. and Rubboli, G. Transcranial magnetic stimulation in epileptic patients: usefulness and safety. Neurology, 1990, 40: 1132-1133. Tenforde, T.S. Interaction of ELF magnetic fields with living matter. In: C. Polk and E. Postow (Eds.), CRC Handbook of Biological Effects of Electromagnetic Fields. CRC Press, Boca Raton, FL, 1986: 198-220. Thomas, S., Merton, W.L. and Boyd, S.G. Pituitary hormones in relation to magnetic stimulation of the brain. J. Neurol. Neurosurg. Psychiat., 1991, 54: 89-90.
215
ELMOCO 91121
M a g n e t i c brain s t i m u l a t i o n a n d brain size: relevance to a n i m a l studies J.D. Weissman b, C.M. Epstein a,b and K.R. Davey a,c a Rehabilitation Research and Development Center, Atlanta VA Medical Center, Decatur, GA (U.S.A.), b Department of Neurology, Emory University School of Medicine, Atlanta, GA (U.S.A.), and c Department of Electrical Engineering, Georgia Institute of Technology, Atlanta, GA (U.S.A.) (Accepted for publication: 30 January 1992)
Summary
Magnetic brain stimulation (MBS) is widely used for the investigation of brain function in man, but there have been only a few reports of its safety in animals. These results were predominantly benign, but the effectiveness of stimulation in animals is unclear. Because the stimulators produced obvious motor effects in humans, or had a comparable peak magnetic field strength, they were assumed to produce comparable electric field intensities and neuronal effects in animal brains. We tested this assumption using 3 stimulus coils of different sizes and design, plus 6 saline-filled spheres that spanned a range of volume from 0.5 to 1800 ml. The induced electric field diminished monotonically with decreasing radius, by factors of 4.7-6.2 at the extremes of size. Comparable results were found using a mathematical model. These results suggest that the efficiency of magnetic stimulation is drastically reduced in smaller brains, and that threshold and safety studies in some animal models may not be valid.
Key words: Magnetic brain stimulation; Animal studies; Electric field; Magnetic field; Safety
Magnetic brain stimulation (MBS) is increasingly used for the investigation of motor pathways in patients and normal subjects (Barker et al. 1986; Schriefer et al. 1987, 1989; Claus et al. 1988; Hallett and Cohen 1989; Miiller et al. 1991). With a limited number of single stimuli, its application is generally considered safe (Barker et al. 1990; Tassinari et al. 1990). Brief trials of MBS in humans have not been associated with EEG (Levy et al. 1990), SPECT (Dressier et al. 1990), neuropsychological (Bridgers and Delaney 1989) or hormonal changes (Thomas et al. 1991). Possible induction of seizures has been reported rarely (H6mberg and Netz 1989). Trains of magnetic stimuli may provoke typical seizures in patients with pre-existing epilepsy (Hufnagel et al. 1990), and early reports of rapid repetitive MBS suggest a greater seizure-evoking ability (Pascual-Leone et al. 1991). In general, however, human safety studies have used limited numbers of subjects and stimuli, and follow-up has been brief. Pathologic examination has been possible only in patients about to undergo brain resection for the treatment of intractable epilepsy (Pascual-Leone et al. 1991). Surprisingly few investigators have described long-term effects of MBS in experimental animals. Small animals can be exposed to more stimuli
Correspondence to: Charles M. Epstein, M.D., Rehabilitation Research and Development Center, Atlanta VA Medical Center, 1670 Clairmont Road, Decatur, GA 30033 (U.S.A.).
than humans, at all stages of development and for a greater fraction of their life span. Histological analysis might detect subtle long-term effects that are difficult to recognize in human patients. But only a few, mostly preliminary, reports of animal safety studies are available for MBS. Although vacuolar changes are described by Matsumiya et al. (1992), others have reported no changes (Mano et al. 1989; Hersh et al. 1990; Kling et al. 1990; Sgro et al. 1990). Kling et al. (1990) felt that the power of their magnetic stimulator would be greater in rats because of the thinner skull, but currently available stimulators may be less effective in animals than in man. The practice of using the peak magnetic field strength as the sole measure of magnetic stimulating power may be misleading. Although responses similar to those of humans have been reported in monkeys (Amassian et al. 1987), dogs (Heckmann et al. 1989) and cats (Fujiki et al. 1990), they have been absent in mice (Hersh et al. 1990) and guinea pigs (R. Gilmore, personal communication). In our experience and elsewhere (from conversations with other investigators) it seems difficult to determine if motor responses to MBS are present in small animals or, if present, are due to activation at a central rather than a peripheral site. Even the body jerking described by Kling et al. (1990) in rats might be due to activation of lower-threshold peripheral nerve or skeletal muscle rather than brain. The apparent differences among species may be due simply to variations in the organization of the motor system. An
216 alternative possibility, however, is that magnetic stimulators are less efficient at inducing current in the confines of a smaller head. We report here studies which indicate that magnetic stimulus coils appropriate for humans produce substantially weaker effects in the much smaller brains of common experimental animals.
Methods
Magnetic stimulator and coils The stimulator and two of the coils, which had a modified "butterfly" configuration, have previously been described (Epstein et al. 1990). The larger and smaller butterfly coils had central segments measuring 10 and 5.3 cm respectively. For these studies we added a circular coil of comparable inductance (10/zhenrys), with an inside diameter of 8.5 cm and outside diameter of 11 cm. Brain models and electric field measurements Theoretical analyses have shown that, for modeling the currents induced during magnetic brain stimulation, the brain and skull can be represented as a homogenous conductor surrounded by an insulator, with boundaries corresponding to the inner table of the skull (Cohen and Cuffin 1991). H~im~il~iinen and Savas (1989) used a similar model for magnetoencephalography, which may be applied to magnetic brain stimulation by invoking the reciprocity theorem. We therefore constructed 6 thin-walled spherical models with internal radii of 0.5-7.5 cm. Corresponding volumes were 0.5-1800 ml, spanning the brain size range from mouse to human. To normalize wall thickness and represent the skull, extra layers of hard rubber insulation were added between some of the spherical models and the stimulus coil. The gap from stimulator to saline boundary was 5 mm for the 4 largest spheres, and 2.5 mm for the two smallest. We tested 3 different conductive media: 0.9% saline (Maccabee et al. 1990), 0.08% normal saline (Geddes and Bourland 1990), and a saturated NaC1 solution approximately equal to 4 N. The spheres were filled entirely with saline solution and a test probe was inserted through a small hole in the top, with the stimulus coil directly beneath the sphere. Measurements were made with a variety of twisted pair (Geddes and Bourland 1990) and concentric (Maccabee et al. 1990) probes, using both bare copper and silver-silver chloride electrodes. Probes were placed tangent to the inner wall of the sphere 2 mm inside the saline boundary. For the butterfly coils, all measurements were made above the middle of the central segment, where the induced electric field was known to be largest (R6sler et al. 1989; L.G. Cohen et al. 1990; Epstein et al. 1990; Maccabee et al. 1990). For the round coil, measurements were made 2 mm from
J.D. WEISSMAN ET AL. the inner wall of the sphere at the point where it contacted the coil. To avoid distortion by current within the probe, the input to a storage oscilloscope was fed through a high impedance unity-gain field-effect transistor (FET) amplifier. Capacitive effects were minimized by a 20 × 20 cm copper screen Faraday shield between the coil and the model head. Throughout the course of the measurements we regularly reversed the direction of the probe to test for polarization effects and turned the probe at right angles to the coil windings to check for effects of capacitance. In the absence of significant eddy currents, the induced electric field should be proportional to the rate of change of the magnetic flux captured by the target, which is in turn a function of its size (Tenforde 1986). To test this hypothesis we constructed a series of round single-turn search loops, with radii matching the positions of the probes within the saline-filled spheres.
Mathematical modeling of induced fields The effect of target size was calculated theoretically with a model of homogenous spheres centered over a round stimulus coil, using elliptical integrals to predict the field produced by the coil and the Gauss-Legendre integration technique (Abramowitz and Stegun 1970; Davey et al. 1991) to determine the flux at the test region. Integrating this field yields the flux. Our model coil had a radius of 5 cm with negligible thickness and carried 1000 A turns with a rise time of 100/xsec. The electric fields were arrived at by integrating the magnetic flux over an area within the volume of interest (at a depth equal to 10% of the radius) using the integral form of Maxwell's equations.
Results
Electric field measurements Preliminary studies showed that the electric field strength at equivalent positions was identical for 0.08% saline, normal saline, and the saturated saline solution. This was the expected result in the absence of significant eddy currents (Tenforde 1986). A high noise level was obtained initially from the 3 smallest spheres and dilute saline. Therefore saturated saline was used with the smaller spheres. Comparable electric field measurements were obtained with copper and silver-silver chloride electrodes of both varieties; polarization effects were absent with the brief pulses and very high impedance amplifier used. However, the bare twistedwire pair was easiest to position in the smaller spheres. Consequently a single twisted-wire pair with a gap of 4 mm was used in one consecutive session to obtain all the measurements of Fig. 1. Reversal of probe direction invariably showed that amplitude was unchanged. Wave shape was identical to that measured with a
MAGNETIC STIMULATION AND BRAIN SIZE
217
search coil in air. With the probe oriented at right angles to the stimulator coil windings, the measured potential was always less than 25% of that recorded in the parallel measurements. This and the fact that the induced voltage followed the rate of change of the coil current excluded substantial capacitive effects. Fig. 1 shows the relationship between the diameters of the 6 test spheres and the potentials measured 2 mm inside the inner wall. For all 3 coils, the potentials fall substantially with the radius of the sphere, by factors of 4.7-6.2 at the extremes of size. These ratios are probably conservative because of residual capacitance, with y-intercepts of approximately 0.1 V / c m . For comparison, brain weights of human (1400 g, Neal and Rand 1936), orangutan (350 g, Neal and Rand 1936) and rabbit (30 g, Crosby and Schnitzlein 1982) were taken from standard sources. The weights for rat and mouse brain were measured in our lab and were 2 and 0.5 g, respectively. Assuming a specific gravity near 1 and spherical shape, equivalent radii were calculated as shown in Fig. 1. Note that for the smallest targets the electric field is about twice as strong with the smaller butterfly coil, which is half the size of the larger one. We also compared measurements made with the sphere centered over the edge of the round coil, tangent to its plane, with measurements taken at the center of the coil. These studies were performed with a 0.25 cm concentric probe. For the 2 smallest spheres, roughly similar in size to the brains of a mouse and a rat, the measured potentials were 30% higher with the target in the center of the coil than at the edge. However, they were still several times lower than for the largest sphere.
__A-- L a r g e
Small Butterfly
.o.
Round
Butterfly
1.20 E
1.00
Spherical
Radius,
2.00
1.50
°E
>
/
1.00
/
0.50
/
f
0.00 0
1
2
3
4
5
6
test region radius cm
Fig. 2. Mathematical prediction of peak electric field at or above the center of a circular stimulus coil, calculated for spherical targets of 0-6 cm radius.
The peak voltages induced in the single-turn circular search coils were highly correlated with the radii of the respective head models, with correlation coefficients r--0.98, 0.95 and 0.99 for the small butterfly, large butterfly and round coil, respectively (all P < 0.01).
Mathematical modeling We calculated the induced fields for the case of targets smaller than the stimulus coil along a line passing through the center of the coil perpendicular to its plane. For radii greater than 5 cm, the target could no longer lie in the center of the coil, but contacted it at all points. Fig. 2 illustrates the calculated results. As expected, the induced electric field approaches zero for a test region of zero radius. The calculated field in the vicinity of 5 cm radius is probably distorted, due to the implicit assumptions of negligible coil thickness and negligible skull thickness. Otherwise the mathematical model shows the same behavior as the empirical measurements in Fig. 1.
cm
0
j 0.80
C
....
~
Discussion
.... ; ~
0.60
_u 0.40 LLI
/
0,20 0.00
'
0
1
::
i r~lt mouse rabbit
'
'
'
~
~
J
J
2
3
4
5
6
7
8
orangutan
human
Fig. 1. Peak electric fields, measured with a twisted-wire probe, using 3 different magnetic stimulus coils and 6 spherical head models of different sizes. Representative brain volumes of 5 species are shown for comparison according to their equivalent radius.
These results indicate that the efficiency of MBS drops substantially when the target becomes smaller than the stimulus coil. Experimentally this relationship holds true both for circular coils, which are the most commonly used, and for butterfly coils. The electric field difference which drives current flow is dependent on the changing magnetic flux through the target. A smaller target will capture a proportionately smaller fraction of the total flux. Even if, within the brain, membrane depolarization depends on the electric field gradient (Reilly 1989; Roth and Basser 1990), the gradient decreases as the target volume decreases. This
218
can be analytically shown with the target centered on the coil axis and implicitly shown with the target offaxis. As the volume of the brain falls from that of a human to that of a mouse, the magnitude of the induced field drops by a ratio of at least 5 to 1. But with magnetic stimulation in humans, the threshold of movement typically occurs around 50% of full output (Amassian and Cracco 1987; Amassian et al. 1988; Ackermann et al. 1991; Miiller et al. 1991). If the neuronal thresholds in different species are comparable, even full output would seem insufficient for effective stimulation in small rodents. Our measurements support the technique chosen by Sgro et al. (1990), who stimulated the rat brain inside the center of a round stimulus coil. This is not possible for the human head and body, whose dimensions are larger than the coil itself. Nonetheless, stimulus efficiency is still lower than for a larger volume. Maccabee et al. (1991) reported analogous results with stimulation of sensory fibers in the fingers. They found that small volume conductors such as the fingers required targeting near the center of a round coil instead of under the windings in order to elicit a sensory action potential. Figs. 1 and 2 suggest a simple rule-of-thumb for magnetic brain stimulation with heads of different sizes: the stimulus coil should be comparable in size to the brain, or smaller for greater focality. But for small animals this principle collides immediately with a fundamental physical difficulty. Smaller coils are subject to progressively greater thermal and mechanical stresses and may fail catastrophically (Fortescue and Bickford 1990; Cohen and Cuffin 1991). Practical coils for small brain stimulation may require different geometries or iron cores and may be a challenge to design. The effects of brain size also may account in part for the decrease in magnetic stimulation threshold during human maturation (Koh and Eyre 1988). Similarly, any comparison of the thresholds for magnetic stimulation in brains of different species must take into account the changing efficacy of stimulation associated with differences in brain size. Meanwhile animal safety studies of MBS remain desirable but limited, and these considerations may help to explain the paucity of reports. Our results raise doubts about the validity of some studies conducted thus far, especially those in which large stimulating coils designed for human studies are applied to small animal brains or tissue cultures in small containers. References Abramowitz, M. and Stegun, C. Handbook of Mathematical Functions, 9th Edition. Dover Publications, 1970: 591-592. Ackermann, H., Scholz, E., Koehler, W. and Dichgans, J. Influence
J.D. WE1SSMAN ET AL. of posture and voluntary background contraction upon compound muscle action potentials from anterior tibial and soleus muscle following transcranial magnetic stimulation. Electroenceph. clin. Neurophysiol., 1991, 81: 71-80. Amassian, V.E. and Cracco, R.Q. Human cerebral cortical responses to contralateral transcranial stimulation. Neurosurgery, 1987, 20: 148-155. Amassian, V.E., Quirk, G.J. and Stewart, M. Magnetic coil versus electrical stimulation of monkey motor cortex. J. Physiol. (Lond.), 1987, 304: lt9P. Amassian, V.E., Bigland-Ritchie, B., Cracco, R.Q. and Maccabee, P.J. Motor unit fields mapped by focal magnetic coil stimulation of human motor cortex. J. Physiol. (Lond.), 1988, 420: 20P. Barker, A.T., Freeston, I.L., Jalinous, R. and Jarratt, J.A. Clinical evaluation of conduction time measurements in central motor pathways using magnetic stimulation of the human brain. Lancet, 1986, i: 1325-1326. Barker, A.T., Freeston, I.L., Jarratt, J.A. and Jalinous, R. Magnetic stimulation of the human nervous system: an introduction and basic principles. In: S. Chokroverty (Ed.), Magnetic Stimulation in Clinical Neurophysiology. Butterworth, Boston, MA, 1990: 55-72. Bridgers, S.L. and Delaney, R.C. Transcranial magnetic stimulation: an assessment of cognitive and other cerebral effects. Neurology, 1989, 39: 417-419. Cadwell, J. Principles of magnetoelectric stimulation. In: S. Chokroverty (Ed.), Magnetic Stimulation in Clinical Neurophysiology. Butterworth, Boston, MA, 1990: 13-32. Claus, D., Harding, A.E., Hess, C.W., Mills, K.R., Murray, N.M. and Thomas, P.K. Central motor conduction in degenerative ataxic disorders: a magnetic stimulation study. J. Neurol. Neurosurg. Psychiat., 1988, 51: 790-795. Cohen, D. and Cuffin, B.N. Developing a more focal magnetic stimulator. Part 1. Some basic principles. J. Clin. Neurophysiol., 1991, 8: 102-111. Cohen, L.G., Roth, B.J., Nilsson, J., Dang, N., Panizza, M., Bandinelli, S., Friauf, W. and Hallett, M. Effects of coil design on delivery of focal magnetic stimulation. Technical considerations. Electroenceph. clin. Neurophysiol., 1990, 75: 350-357. Crosby, E.C. and Schnitzlein, H.N. Correlative Neuroanatomy o1 the Vertebrate Telencephalon. McMillan Publishing Co., New York, 1982:414 pp. Davey, K.R., Chin, H.C. and Epstein, C.M. Prediction of magnetically induced electric fields in biological tissue. IEEE Trans. Biomed. Eng., 1991, 38: 418-426. Dressier, D., Voth, E., Feldmann, M. and Benecke, R. Safety aspects of transcranial brain stimulation in man tested by single photon emission computed tomography. Neurosci. Lett., 1990, 119: 153155. Epstein, C.M., Schwartzenberg, D.G., Davey, K.R. and Sudderth, D.B. Localizing the site of magnetic stimulation in humans. Neurology, 1990, 40: 666-670. Fortescue, P. and Bickford, R. The design of electromagnetic stimulators. In: S. Chokroverty (Ed.), Magnetic Stimulation in Clinical Neurophysiology. Butterworth, Boston, MA, 1990: 45-53. Fujiki, M., Isono, M. and Hori, S. Spinal and muscle motor evoked potentials following magnetic stimulation in cats. Neurol. Med. Chir. (Tokyo), 1990, 30: 234-241. Geddes, L.A. and Bourland, J.D. Fundamentals of eddy-current (magnetic) stimulation. In: S. Chokroverty (Ed.), Magnetic Stimulation in Clinical Neurophysiology. Butterworth, Boston, MA, 1990: 33-43. Hallett, M. and Cohen, L.G. Magnetism, a new method for stimulation of nerve and brain. JAMA, 1989, 262: 538-541. H~im~il~iinen, M.S. and Savas, J. Realistic conductivity geometry model of the human head for interpretation of neuromagr~etic data. IEEE Trans. Biomed. Eng., 1989, 36: 165-171.
MAGNETIC STIMULATION AND BRAIN SIZE Heckmann, R., Hess, C.W., Hogg, H.P., Ludin, H.P. and Wiestner, T. Transcranial magnetic stimulation of the motor cortex and percutaneous magnetic stimulation of the peripheral nervous structures in the dog. Schweiz. Arch. Tierheilk., 1989, 131: 341350. Hersh, S.M., Green, R.C., Weissman, J.D., Rees, H.C., Davey, K.R., Epstein, C.M. and Bakay, R.A.E. Biological consequences of transcranial magnetic stimulation in the mouse. Abstract presented at Society for Neuroscience 1990 Annual Meeting, St. Louis, MO, November 1, 1990. H6mberg, V. and Netz, J. Generalized seizures induced by transcranial magnetic stimulation of motor cortex. Lancet, 1989, ii: 1223. Hufnagel, A., Elger, C.E., Durwen, H.F., Boker, D.K. and Entzian, W. Activation of the epileptic focus by transcranial magnetic stimulation of the human brain. Ann. Neurol., 1990, 27: 49-60. K.ling, J.W., Yarita, M., Yamamoto, T. and Matsumiya, Y. Memory for conditioned taste aversions is diminished by transcranial magnetic stimulation. Physiol. Behav., 1990, 48: 713-717. Koh, T.H. and Eyre, J.A. Maturation of corticospinal tracts assessed by electromagnetic stimulation of the motor cortex. Arch. Dis. Childh., 1988, 63: 1347-1352. Levy, W.J., Oro, J., Tucker, D. and Haghighi, S. Safety studies of electrical and magnetic stimulation for the production of motor evoked potentials. In: S. Chokroverty (Ed.), Magnetic Stimulation in Clinical Neurophysiology. Butterworth, Boston, MA, 1990: 165-172. Maccabee, P.J., Eberle, L.P., Amassian, V.E., Cracco, R.Q., Rudell, A. and Jayachandra, M. Spatial distribution of the electric field induced in volume by round and figure 8 magnetic coils: relevance to activation of sensory nerve fibers. Electroenceph. clin. Neurophysiol., 1990, 76: 131-141. Maccabee, P.J., Amassian, V.E., Cracco, R.Q., Eberle, L.P. and Rudell, A.P. Mechanisms of peripheral nervous system stimulation using the magnetic coil. In: W.J. Levy, R.Q. Cracco, A.T. Barker and J.C. RothweU (Eds.), Magnetic Motor Stimulation. Electroenceph. clin. Neurophysiol., Suppl. 43. Elsevier, Amsterdam, 1991: 340-357. Mano, Y., Funakawa, I., Nakamuro, T., Takayanagi, T. and Matsui, K. The kinesiological, chemical and pathological analysis in pulsed magnetic stimulation to the brain. Rinsho Shinkeigaku, 1989, 29: 982-988. Matsumiya, Y., Yamamota, T., Yarita, M., Miyauchi, S. and Kling, J.W. Cortical damage by magnetic stimulation. J. Clin. Neurophysiol., 1992, in press.
219 Miiller, K., H6mberg, V. and Lenard, H.-G. Magnetic stimulation of motor cortex and nerve roots in children. Maturation of corticomotoneuronal projections. Electroenceph. clin. Neurophysiol., 1991, 81: 63-70. Neal, H.V. and Rand, H.W. Comparative Anatomy. Blaki~ton Publishing Co., Philadelphia, PA, 1936:505 pp. Pascual-Leone, A., Gates, J.R. and Dhuna, A. Induction of speech arrest and counting errors with rapid transcranial magnetic stimulation. Neurology, 1991, 41: 697-702. Reilly, J.P. Peripheral nerve stimulation by induced electric currents: exposure to time-varying magnetic fields. Med. Biol. Eng. Comput., 1989, 27: 101-110. Roth, B.J. and Basser, P.J. A model of the stimulation of nerve fiber by electromagnetic induction. IEEE Trans. Biomed. Eng., 1990, 37: 588-597. Roth, B.J., Saypol, J.M., Hallett, M. and Cohen, L.G. A theoretical calculation of the electrical field induced in the cortex during magnetic stimulation. Electroenceph. clin. Neurophysiol., 1991, 81: 47-56. R6sler, K.M., Hess, C.W., Heckmann, R. and Ludin, H.P. Significance of shape and size of the stimulating coil in magnetic stimulation of the human motor cortex. Neurosci. Lett., 1989, 100: 347-352. Schriefer, T.N., Mills, K.R., Murray, N.M.F. and Hess, C.W. Magnetic brain stimulation in functional weakness. Muscle Nerve, 1987, 10:463 (abstract). Schriefer, T.N., Hess, C.W., Mills, K.R. and Murray, N.M. Central motor conduction studies in motor neurone disease using magnetic brain stimulation. Electroenceph. clin. Neurophysiol., 1989, 74: 431-437. Sgro, J.A., Stanton, P.C., Emerson, R.G. and Ghatak, N.R. Repetitive high magnetic field stimulation: the effect on rat brain. Electroenceph. clin. Neurophysiol., 1991, 78: 24P. Tassinari, C.A., Michelucci, R., Forti, A., Plasmati, R., Troni, W., Salvi, F., Blanco, M. and Rubboli, G. Transcranial magnetic stimulation in epileptic patients: usefulness and safety. Neurology, 1990, 40: 1132-1133. Tenforde, T.S. Interaction of ELF magnetic fields with living matter. In: C. Polk and E. Postow (Eds.), CRC Handbook of Biological Effects of Electromagnetic Fields. CRC Press, Boca Raton, FL, 1986: 198-220. Thomas, S., Merton, W.L. and Boyd, S.G. Pituitary hormones in relation to magnetic stimulation of the brain. J. Neurol. Neurosurg. Psychiat., 1991, 54: 89-90.
