The specific location on the magnetic stimulation (MS) coil that may correspond to the area of nerve depolarization has not been determined. In order to localize such an area, MS with 9-cm and 5-cm diameter coils was compared with conventional percutaneous electric stimulation (ES). On the 9-cm coil the distribution of points of nerve depolarization corresponded to that quarter of the coil which was placed over and parallel to the median nerve, whereas on the 5-cm coil, this area also extended outside the coil. The points of median nerve depolarization with MS were distributed over a distance of 7 cm on the stimulator head and was nearly identical for the 2 coil sizes at the wrist and elbow. Ulnar nerve costimulation was less frequent with the smaller coil at the wrist. A calculated reference point on the coil is suggested for more accurate NCV determinations. Key words: magnetic stimulation median nerve electromagnetics MUSCLE 81 NERVE 15:711-715 1992
LOCALlZATlON OF NERVE DEPOLARlZATlON WITH IB R. ODDERSON, MD, PhD, and EUGEN M. HALAR, MD
Magnetic stimulation (MS) is a relatively new technique for peripheral nerve stimulation and holds potential applications that have not been possible with conventional ES. MS is considered less painful and may be useful for children and adults who cannot tolerate percutaneous ES. MS may also offer a technique for obtaining nerve conduction studies in patients with casts and be of use in edematous tissues without requiring invasive techniques. A major problem with MS, however, is the large size of both the coil and the induced electromagnetic field, posing significant difficulties in determining the pont of nerve stimulation for NCV calculations. 1,274s7 This was verified by a study in which MS was used to stimulate 2 peripheral nerves simultaneously .3 T h e specific area on the MS coil that may correspond to nerve depolarization has not been determined. In order
From the Department of Rehabilitation Medicine (Dr. Odderson), and University of Washington School of Medicine Seattle, Washington and Veterans Administration Medical Center Seattle, Washington (Dr. Halar). Presented at the combined Annual Meetings of the American Academy of Physical Medicine and Rehabilitation and the American Congress of Rehabilitation Medicine, Orlando, Florida, November, 1989. Acknowedgment: The authors thank Drs. Walter C. Stolov and Lawrence R. Robinson for their careful review of the manuscript Address reprint requests to Ib R. Odderson. MD. PhD, Department of Rehabilitation Medicine, University of Washington School of Medicine, Seattle, WA 98195. Accepted for publication September 22, 1991 CCC 0148-639)(/92/060711-05$04.00 0 1992 John Wiley & Sons, Inc.
Localization of Nerve Depolarization
to localize an area of nerve depolarization, various positions and orientations of the coil have been investigated and compared to standard technique~.~,~ Maccabee et al.4 studied the effects of various magnetic coil orientations and specific positions of the coil in relation to the median nerve. They found that the induced voltages and CMAPs were largest when the MS-induced field was oriented parallel to the long axis of a metallic wire and to the median nerve, respectively. Furthermore, with supramaximal magnetic stirnulation the area of impulse origin of the nerve propagation varied 2 to 15 mm from the midpoint of the contacting edge of the MS coil. Evans et a1.* found that the best position of the stimulator head for nerve activation was against the skin and along the course of the nerve, 1 to 2 cm within the circle of the contained stimulator coil. However, no consistent single point of depolarization could be found. Both studies confirmed that the actual point of nerve stimulation varies widely with small shifts in the stimulator position. In a recent study, Roth and Basser’ have, by means of a model, defined the expected region of nerve depolarization. They found, for a 5-cm diameter coil, this region was in the posterior quarter of the coil overlying the nerve. However, if stronger stimuli are applied, several centimeters of the nerve length is simultaneously brought above thresold and, therefore, it is difficult to determine the exact location of stimulation. The
MUSCLE & NERVE
same authors8 found, in a mathematical model, that the coil placed above and oriented perpendicular to the arm produced a small electric field with rapid fall off below the skin. Positioning the coil above but parallel to the arm produced a larger electric field that fell off more slowly with depth. Cohen et al.' measured the magnetic fields induced by several different coils and calculated their electric fields. For all the coils excpet the butterfly-shaped coil, the largest magnetic field was at the circumference of the coil and decreased toward its center. Consequently, there was no focal delivery of the stimulus with a single coil design stimulator. Olney et al? found, in a small number of subjects, that the butterfly-shaped coil provided a stimulus that was slightly more focal than the round coil at the wrist, but not as narrow a field as achieved by ES. Considering the potential new applications of MS the purpose of this study was to determine the location on the MS coil that may correspond to the area of nerve depolarization. By placing only one quarter of the coil over and parallel to the tested nerve, we intended to diminish the area of coil stimulation so that the narrow area of nerve depolarization could be identified. T h e objectives of the study were: (1) to localize the area on the MS coil that best corresponded to the point of nerve depolarization using ES as the standard; (2) to assess the importance of MS coil size; and (3) the effect of anatomical site (proximal vs. distal) for localizing the area of nerve depolarization on the coil; and (4) to determine the amount of ulnar nerve costimulation when stimulating the median nerve using small and large magnetic coils.
at 100% output, and has a stimulus pulse rise time of 70 ps, pulse width of 70 ps, and a decay time of' 70 ps. T h e median nerve was stimulated at the wrist and at the elbow with both ES and MS; a 5-cm coil was used for stimulation of both arms while a 9-cm coil was only used on the left arm. T h e ulnar nerve was stimulated approximately 1 cm proximal to the proximal wrist crease and below the elbow with ES only. Simultaneous tracings were made from the abductor pollicis brevis (APB) and abductor digiti rninimi (ADM) on a 2-channel EMG recorder (Viking, Nicolet). Nerve conduction velocities were determined with ES for the median and ulnar nerves in the forearm. T h e stimulating coil was placed on the skin with one-quarter of the coil over and parallel to the median nerve and the remaining portion of the coil directed away from the ulnar nerve (Fig. 1). No part of the coil was overlying the ulnar nerve in order to minimize ulnar nerve costimulation. T h e position was adjusted to obtain a clear take off on the evoked response and minimum stimulation of the ulnar nerve. That coil position was carefully traced on the skin by a permanent marker pen. The point of nerve depolarization
Thirty-four median and ulnar nerves in 22 healthy subjects, ages 26 to 41 years, were studied after obtaining informed consent in accordance with the Institutional Human Subjects Committee. The groups of subjects tested with each coil were not identical although some were tested with both coils. The testing was carried out with skin temperatures all above 30.0" C. Percutaneous ES was performed with square wave stimuli of 0.1 to 0.2 ms duration and u p to 80 mA supramaximal current (Viking, Nicolet Instrument Corp. Madison, WI). T h e MS was performed with 5-cm and 9-cm coils, with right angles at the tip of the stirnulator head (Focalpoint, Cadwell Laboratories Inc., Kennewick, WA) and magnetic stimulation at 70% to 100% output (MES-10, Cadwell). This magnetic stimulator produces a maximal field flux of 2.2 T
Localization of Nerve Depolarization
FIGURE 1. Position of magnetic stimulator head at the wrist
MUSCLE & NERVE
for MS was then calculated by multiplying the difference in motor latencies of the 2 techniques with the forearm ES nerve conduction vel~city.',~ This calculated distance from the point of nerve depolarization by ES indicates the MS point of depolarization which was marked on the skin and verified by ES. If the ES verification of MS latencies did not agree, ES was continued until matching latencies were found and this ES point was then used as the point of nerve depolarization with MS. A transparent film sheet with an outline of the MS coil shape was then placed over the tracing of the MS coil on the skin, and the calculated point of nerve depolarization recorded on the transparent film sheet. When ulnar nerve costimulation occurred with MS, the thenar CMAP response was corrected by subtracting a fraction representing the degree of ulnar nerve contribution to the thenar CMAP.' The maximum thenar response of the ulnar nerve at the wrist obtained with ES was multiplied by the ratio of hypothenar CMAP, as obtained by MS of the median nerve to the hypothenar CMAP by supramaximal ES of the ulnar nerve and subtracted from the thenar response. The thenar response was corrected as follows when ulnar nerve costimulation occurred:
FIGURE 2. The calculated points on the coils at which median nerve depolarization occurred.
one-quarter of the coil that was placed over and parallel to the median nerve (Fig. 2). The points of nerve depolarization were located in the area measuring 7.0 x 3.2 cm, and were essentially similar in size to the 9-cm coil. A comparison of the points of nerve depolarization obtained at the wrist versus elbow is illustrated in Figure 3. The area of nerve depolarization was, in most cases, located to that part of the stimulator head which was overlying the nerve when stimulating at the elbow, whereas at the wrist that area extended outside the perimeters of ThenarCMAPc,,,ec,,d = ThenarCMAPMs,,,d the stimulator head. Comparison of left- versus - HypothenarCMAPMs,,d right-sided stimulation did not demonstrate any HypothenarCMAPESuln xThenarCMAPEsdn difference in location or size of the area of nerve . _ depolarization. In 2 subjects, the points of nerve Paired t-tests were used to evaluate differences bedepolarization with MS were located to an anatween amplitudes for the 2 techniques. tomical site of the forearm, and small movements of the coil did not change this location. RESULTS Area of Stimulation on the Coil
Nine-Centimeter Coil. Median nerve stimulation with the 9-cm coil at the wrist and elbow showed large intersubject variations (Fig. 2). The points of nerve depolarization were located to that quarter of the coil that was placed over and parallel to the median nerve. The points of nerve depolarization were located within the area on the coil measuring approximately 6.0 x 2.8 cm in size. The points of nerve depolarization, as traced on the coil when stimulating at the elbow and wrist, were similar. However, the points of nerve depolarization were located within a smaller area and closer to the tip of the coil when stimulating at the elbow. Five-Centimeter Coil. The area of nerve depolarization for the 5-cm coil extended beyond that
Localization of Nerve Depolarization
The MS-corrected CMAPs were an average of 10.9% smaller than the uncorrected amplitudes ( P < 0.001). T h e corrected CMAPs were significantly smaller for the 5-cm coil than the ES amplitudes at the wrist (average 18.7% smaller; MS: 6.0 k 2.9 vs. ES: 7.7 k 2.7 mA, P < 0.05) and elbow (average 23.8% smaller; MS: 5.7 2 2.8 vs. ES: 7.5 rt 2.6 mA, P < 0.001). Similarly, the 9-cm coil-corrected CMAPs at the wrist were significantly smaller than for ES (average 3.1 31.8% smaller; MS: 7.2 & 3.8 vs. ES: 10.0 mA, P < 0.001). At the elbow, the corrected CMAPs were not statistically different from ES, although they were 18.0% smaller (MS: 8.3 4.7 vs. ES: 9.5 2 2.9 mA). The MS uncorrected CMAPs were an average of 12.1% smaller than the ES amplitudes ( P < 0.001).
MUSCLE & NERVE
nerve depolarization on the MS coil determined in our study was in good agreement with the theoretical regions of maximal nerve depolarization calculated by Koth and B a ~ s e rTheir .~ region of maximal depolarization was centered approximately 4.5 cm posterior to the front of the coil compared with 3.8 cm from the tip of the stimulator head in our study. Although, Roth and Baser7 used a round coil, versus the pointed coil in our study, the findings appear comparable, because the configuration of the electric field for two coils were similar for that part of the coils.' The larger area of nerve depolarization in our study, when compared to the area calculated by Maccabee et a ~ , ~ appeared to be related to the position of the coil in respect to the nerve. The plane of the coil in Maccabee's study was at a 135" to 180" angle to the surface of the forearm. This perpendicular position of the coil, according to Roth et al.,' results in a localized but weak electric field. Placing the coil plane parallel to the nerve and to the skin surface, as in our study, produces a larger electric field, which presumably falls off more slowly with depth of tissue.8 The most focused stimulation and electric field appeared to be with the butterfly-type coil. When comparing proximal (elbow) versus distal (wrist) stimulation sites, the areas of nerve depolarization were smaller with stimulation at the elbow. This is likely due to the more focused induced magnetic field at greater tissue depth.' I t was seen for both coil sizes. Other factors which may play a role in the variability of localizing the nerve depolarization are different nerve membrane properties, thresholds, and tissue conductance proximally versus distally, as suggested by Roth and B a ~ s e r With . ~ improved focusing at the elbow, the greatest applicability for peripheral nerve stimulation appears to be more proximally.2 The corrected CMAP amplitudes were an average of 21.7% smaller with MS than ES due in part to submaximal output of the magnetic stimulator in order to minimize ulnar nerve costimulation and volume conduction. The smaller amplitudes with MS are a potential source of error in latency determination and the MS latencies may thus have been minimally overestimated. However, a small error in determining each point of nerve depolarization with MS would not have altered the size of the proportionately large area of points of nerve depolarization on the MS coil. The conclusions of this article would not have been altered if supramaximal amplitudes could have been obtained with MS. Smaller MS amplitudes 135
FIGURE 3. Upper figures: Cornparison of the sites of median nerve depolarization at the wrist versus elbow for the 5-cm coil. Results from both arms are shown. Lower figures: Comparison of the sites of median nerve depolarizationin the left versus right arm for the 5-cm coil.
Costimulation of the ulnar nerve with MS of the median nerve was greater at the elbow than at the wrist. For the 5-cm and 9-cm coils, 65% and 50% of the ulnar nerves, respectively, were costimulated at the elbow. At the wrist, 15% and 43% of the ulnar nerves were costimulated using the 5-cm and 9-cm coils, respectively. Ulnar Nerve Costimulation.
The sites on the MS coil corresponding to the point of nerve depolarization showed great interand intrasubject variability. The distribution of calculated points of nerve depolarization were confined to a specific area that appeared independent of the size of the coil. For the smaller 5-cm coil, this area of nerve depolarization extended beyond the boundaries of the coil. T h e area of nerve depolarization appeared to be less dependent on the coil size than on the magnetic stimulator output and induced field-flux. The area of
Localization of Nerve Depolarization
MUSCLE & NERVE
were also found by Evans et al.' and Olney et al.5 when the round-shaped coil was used. MS amplitudes of comparable size to ES were only found by Ravnborg et al.' when 100% magnetic stimulator output was used, and no attempts were made to eliniinate costimulation or use corrected amplitudes. Consequently, these reported6 amplitudes with MS were likely overestimated and the result of median nerve costimulation, volume conduction, or anomalous innervation in the forearm, and thus not comparable to ES of a single nerve. Olney et al.,' on the other hand, found that supramaximal stimulation could only be obtained with the butterfly coil. Consequently, the MS techique may not be accurate for measuring amplitudes in the distal limbs, but possibly useful for nerve studies when other methods are not feasible. Ulnar nerve costimulation with MS of the median nerve was least frequent with the 5-cm coil at the wrist, and occurred in 15% of the tested median nerves. Evans et al.' also found ulnar nerve costimulation to be a problem. They were unable to achieve supramaximal stimulation of the median nerve without concomitant activation of the ulnar nerve. No difference in localization of nerve depolarization was observed if the current flow was directed clockwise or counterclockwise. This observation was also previously r e p ~ r t e d . ' . ~ l o improve the accuracy of NCV using MS, a variety of coil nerve positions have been used with unsatisfactory r e ~ u l t s . " ~In order to minimize these variables, the magnetic stimulation techniques for peripheral nerves should be standardized fbr each nerve at each anatomical stimulation site. We recommend the stimulation technique used in this study, whereby one distal quarter of the coil was placed over and parallel to the nerve to be stimulated. The remaining part of the coil is then directed away from the adjacent nerve to eliminate costimulation. This position has proven to be as accurate as ES for the 9-cm coil at the elbow. This is similar to the tangential orientation described by M a ~ c a b e e . ~ Our study has determined the distribution points of nerve depolarization on the MS coil. By selecting the center of these points, it is possible to construct a reference point (stimulator point) for each coil. This reference point is located 3.8 cm from the tip of the stimulator head for both coils. The application of such a reference point may improve the accuracy of NCV determinations when MS is done with round and pointed coils.
Localization of Nerve Depolarization
In summary, although magnetic stimulators are relatively large, with proper positioning over the nerve, the depolarization of the nerve occurs over a limited area below the coil. The area of nerve stimulation on the MS coil is nearly identical in size for the 2 cod sizes tested. Costimulation of the ulnar nerve with MS of the median nerve occurs frequently, but is less frequent with the smaller coil. Determination of a reference point for nerve depolarization on the coil could possibly improve the accuracy of MS. T h e proximal stimulation of periperal nerves offer the most useful application of MS. CONCLUSION
With specific positioning of the magnetic coil over the nerve to be stimulated, the error in determining the conduction latency and evoked potential amplitude is too large for routine clinical use. A calculated reference point for each coil, as determined in this study for peripheral nerve stimulation, is suggested as a new possibility for accurate NCV determination. Further studies are needed to justify its use in routine studies and in restricted clinical studies where deep nerve stimulation is needed.
REFERENCES 1. Cohen LC:, Roth B J , Nilsson J , Dang N, Panizza M , Band-
inelli S, Friauf W, Hallett M: Effects of coil design on delivery of focal magnetic stiniulation. rI'echr~icalconsiderations. Eleclrorncrphalogr Neurophysiol 1990;7 5 :350 - 357. 2. Evans HA, Litchy WJ, Daube JR: T h e utility of magnetic stimulation for routine peripheral nerve conduction studics. M?&c N e m ~1988;11:1074-1078. 3 . Goininak S, Cros DP: Magnetic stimulation F waves. Arch Phys Mcd Rehabil 1989;70:A-60. 4. Maccabee PJ, Amassian VE, Cracco RQ, Cadwell J A : A n analysis of peripheral niotor nerve stimulation in humans using the magnetic coil. Electroencephalogr Clin Nrurophysiol 1988;70:524-533. 5. Olney KK, So Y T , Goodin DS, Aminot'f MJ: A comparison of- magnetic- a n d electrical stimulation of peripheral nerves. Muscle Nc~rve1990;13:957-963. 6. Ravnborg M, Blinkenberg M , Dahl K: Significance of magnetic coil position in peripheral motor nerve stimulation. Murcle N w 7 ~1990;13:681-686. 7. Koth BJ, Basser PI: A model of the stimulation of a nerve fiber by electromagnetic induction. I E E E Trans Riornrd Eng 1990;37:588-597. 8. Roth BJ, &hen I S , Hallett M, Friauf W, Hsser PJ: A theoretical calculation of the electric field induced by magnetic stimulation of a peripheral nerve. Musclr Nerve 1990; 13934-741.
MUSCLE & NERVE