\98

Transcranial Magnetic Stimulation and Transcranial DirectCurrent Stimulation (Supplements to Clinical Neurophysiology. Vol. 56)

Editors: W. Paulus, F. Tergau, M.A. Nitsche, r.c. Rothwell, U. Ziemann, M. Hallett© 2003 Elsevier Science B.V. All rights reserved

Chapter 19

Motor and phosphene thresholds: consequences of corticalanisotropy

Thomas Kammer-v", Sandra Becka, Klaas Puls", Claire Roether andAxel Thielscher'

a Department of Neurobiology, Max Planck Institute for Biological Cybernetics, Tubingen (Germany)b Department of Neurology, University of Tiibingen, Tiibingen (Germany)

C Department Psychiatry Ill, University of Ulm, Ulm (Germany)

1. Introduction

Excitability of cortical areas assessed by TMSdepends on the induced current direction. This hasbeen extensively demonstrated for the motor cortex.In case of monophasic pulses currents passing theprecentral gyrus from back to front are more efficientthan currents in the reverse direction (Brasil-Netoet al., 1992; Mills et al., 1992; Niehaus et al., 2000;Di Lazzaro et al., 2001; Kammer et al., 2oo1b). Theoptimal current orientation is perpendicular to thecentral sulcus (Brasil-Neto et al., 1992; Mills et aI.,1992; Nithi and Mills, 2000). In recent years, it wasfound that other cortical areas also show thisanisotropic response behavior when stimulated withTMS. In the visual system, early reports suggested apreferential current direction from lateral to medial,both in the induction of phosphenes (Meyer et al.,1991) and for the visual extinction effect (Amassianet al., 1994). Our group confirmed this anisotropy

* Correspondence to: Dr. Thomas Kammer, Departmentof Neurology, University Hospital of Tubingen, Hoppe-Seyler-Str, 3, 0-72076 Tubingen, Germany.Tel: +49 7071 29 80469; Fax: +49 7071 29 5326;E-mail: thomas.kammer@uni-tuebingen.de

with the first phosphene threshold study systemati-cally varying current directions (Kammer et al.,2oo1a). Lowest phosphene thresholds were obtainedwith latero-medial currents in both hemispheres,highest threshold with medio-lateral currents. Verticaloriented currents yielded threshold values betweenboth extremes. For a third region, the prefrontalcortex, a current direction preference has beendemonstrated using a behavioral task (Hill et al.,2000).

In the present study we directly compare thecurrent direction effect of monophasic pulses onthresholds in the visual and motor system using twodifferent stimulator types. With a frameless stereo-tactic positioning system, we further demonstrate adependency of phosphene maps from the visualcortex on current direction.

2. Methods

Six healthy subjects (age 21 to 37 years, 4 males, 2females) participated in the study after giving theirwritten informed consent. The study was approvedby the local internal review board of the MedicalFaculty, University of Tiibingen.

Two stimulators were used: the Medtronic-DantecMagpro (Skovlunde, Denmark) in the monophasicmode (maximal rate 0.33 Hz) and the Magstim 200(Whitland, Dyfed, Wales, maximal rate 0.2 Hz). Bothstimulators were fitted with their standard figure-of-8 coil. Coil position in relation to the head wasmonitored and registered continuously in all sixdegrees of freedom - three translational and threerotational - with a custom-made positioning system(Kammer et al., 2oo1b).

Data of phosphene thresholds presented here weretaken from Kammer et aI. (2001a), left hemisphere.They have been measured with the method ofconstant stimuli. Subjects had to report the presenceor absence of a phosphene at stimulation with 10different intensities (step size 2%) tested 10 times ina randomized order. Threshold was then calculatedusing a sigmoidal Boltzmann fit to the data. Threeindependent measurements with 100 stimuli eachwere averaged. For more details, see Kammer et al.(2001a).

Resting motor thresholds were measured in theright abductor pollicis brevis muscle as describedin Kammer et aI. (2oolb). The threshold criterionwas 50!JV peak-to-peak amplitude. Ten trials wereperformed at each stimulator intensity, varied in stepsof I%. Thresholds were calculated using a sigmoidalBoltzmann fit.

In order to compare threshold data from both stim-ulators, they were normalized with respect to ~Emax'

the maximal energy stored in the stimulator (cf.Barker et al., 1991; Kammer et al., 2ootb). Emax ofMedtronic-Dantec is 300 J, Emax of Magstim 200 is720J.

For phosphene mapping, subjects looked at afixation point in the middle of the screen witha background intensity of 0.3 cd/rn-, They observedphosphenes induced by a single TMS pulse and drewthe contours of the observed phosphene directly onthe screen using a mechanical digitizing armprogrammed as a drawing device. Traces of thedrawing appeared as white lines on the screen.Subjects could freely release stimulation pulses inorder to compare the drawings with the perceivedphosphene and if necessary correct the drawing.

199

Finally the drawing was saved on the computertogether with the exact stimulation site of the coil asmeasured by the positioning system. The drawingswere off-line classified according to the visualhemifield the phosphenes appeared in right (coded inred), left (blue) or bilateral (yellow). A 3D mesh ofthe cortical surface was obtained with the softwareBrainVoyager (Brain Innovation, Maastricht, TheNetherlands) based on a segmentation of the whitematter in a Tl-weighted anatomical scan of thesubject. Stored coil positions (midpoint of the figure-of-8 coil, peak electric field) were projected on the3D mesh of the individual subject's cortical surface.They were colored according to the phosphene hemi-field classification.

3. Results

Mean normalized threshold values from the left visualand motor cortex are shown in Fig. I. They weresubjected to a three-way analysis of variance(ANOVA) with factors of threshold type (phosphenevs. motor), stimulator type (Medtronic-Dantec vs.Magstim), and current direction (latero-medial vs,medio-lateral for phosphene threshold, antero-poste-rior vs. postero-anterior for motor threshold).Significant main effects of stimulator type(F(1,5) =144.6; p < 0.0001) and current direction(F(1,5) =113.7; p < 0.0002) were obtained. Nosignificant difference was found for the factorthreshold type (F(I,5) =1.54; p =0.27), as well asno significant interaction. Medtronic-Dantec wasmore efficient compared to Magstim. The ratio ofMagstimlMedtronic-Dantec thresholds in the visualsystem was 1.26; in the motor system the ratio was1.32. The mean threshold ratio for the differentcurrent directions was 1.19 in the visual system(medio-lateral/latero-medial) and 1.32 in the motorsystem (antero-posterior/postero-anterior).

Correlation of phosphene and motor thresholds forthe individual subjects is depicted in Fig. 2 for eachof the four measurement types (two stimulators, twocurrent directions). Linear regressions revealed nosystematic dependency of phosphene and motorthresholds in the individual subjects.

200

0.L---r-----r----r-------,.-----

In Fig. 3 a map of perceived phosphenes in depen-dence of the coil position is shown. Horizontallyoriented stimulation pulses at a constant stimulusintensity were applied at several sites of the occipitalcortex. The pins indicate the midpoint of the coilinducing a certain phosphene coded by the color ofthe pin. At the lateral stimulation sites over onehemisphere the evoked phosphenes appeared in thecontralateral visual hemifield (red or blue pins). Moremedial stimulation sites resulted in the perception ofphosphenes in both visual hemifields (yellow pins)indicating the supra-threshold stimulation of bothhemispheres. The borders between unilateral andbilateral stimulation depend on the induced currentdirection which is indicated by the arrows. They werenot found in symmetry to the interhemispheric cleftbut were shifted in the direction of the inducedcurrent. In the depicted example the amount of shiftwas 12 mm. The shift could be observed with bothstimulators. Magstim 200 as well as Medtronic-Dantec in the monophasic mode. It was confirmed infour subjects investigated. ranging from 7-15 mm

Phosphene

......... Magstim

......... Medtronlc-Dantec

Motor

20

antero-posterior poslero-anlerior medlo-lateral latero-medlal

Induced current direction

Fig. I. Comparison of motor and phosphene thresholds(mean ± SO) measured in the left hemisphere in sixsubjects. Motor thresholds in the right abductor policisbrevis muscle were measured with the coil over the leftmotor hot spot. The coil handle was oriented perpendicularto the central sulcus. Phosphene thresholds were measuredover the left occipital pole with the coil handle orientedhorizontally. Notice that thresholds were normalized withrespect to the maximal energy stored in the stimulator. For

ANDYA. see text.

20 • MagsUm optimal0 Magstim nonoptimal

18 A Medtronic-Dantec optimal/). Medtronic-Dantec nonoptimal

160

~0

14 -,0".

~ 12 • 0···.5'•

.r:::. A A-~j 10

r8

6 •6 8 10 12 14 16 18 20

Phosphene threshold

Fig. 2. Correlation of threshold values between motor andvisual cortex. The normalized individual values of sixsubjects x four conditions are plotted. The two differentstimulator types are coded as dots (Magstim) and triangles(Medtronic-Dantec). Induced current directions are indicatedby filled (optimal direction, latero-medial or postero-anterior) and open (non optimal direction: medio-Iateral orantero-posterior) symbols. Linear regression lines are solid(optimal direction) or stippled (non optimal direction). None

of the linear regressions was statistically significant.

Fig. 3. Phosphene map in dependence of induced currentdirection. The upper and lower images are two differentviews on the occipital pole of subject KP, showing thesame measurement. Coil positions (midpoint of the figure-of-8 coil) are depicted as colored pins. The color indicatesthe visual hemifield where the subject perceived aphosphene. Red: right hemifield, blue: left hemifield,yellow: both visual hemifields. The induced currentdirection is indicated by the large arrows. Stimulationparameters: Medtronic-Dantec, monophasic, 80% intensity.

(data not shown). It did neither depend on thestimulation strength nor on the horizontal position ofthe coil but could be observed within a large areaover the occipital pole.

4. Discussion

Excitability of the visual cortex depends on currentdirection, similar to excitability of the motor cortex.In case of monophasic pulses in both hemispheres atthe occipital pole latero-medial currents are moreefficient compared to medio-lateral currents (Kammeret al., 2001a). The data presented here extend ourprevious results by direct comparison of phosphenethresholds and motor thresholds in the same subjects.The mean of motor thresholds in six subjectsconfirms the pattern reported in a previous study(Kammer et al., 200lb), i.e, lower thresholds forpostero-anterior oriented currents compared toantero-posterior currents. Furthermore, normalizationof threshold values with respect to the maximal storedenergy (Barker et aI., 1991; Kammer et al., 2001b)reveals that the Medtronic-Dantec stimulator trans-fers stimulation energy more efficient to the braincompared to the Magstim stimulator. The mainreason for this difference is the coil geometry of thetwo figure-of-8 coils, as shown in a modelingapproach (Thielscher, 2(02). In case of Medtronic-Dantec the two circular windings are bent to eachother forming an angle of 1400

, whereas in case ofMagstim the two windings are placed in a plane thusforming an angle of 180.

In our sample of six subjects we did not observea significant correlation between motor and visualthresholds. This finding confirms previous observa-tions (Stewart et al., 2001; Boroojerdi et al., 2(02)that excitability of the two areas differ within asubject. However, we cannot exclude that theseresults are based on variations of the coil-cortexdistance within a subject (Kozel et al., 2000;McConnell et al., 2(01). Further studies combiningcareful threshold measurement and a model of fieldstrength in dependence of the coil-cortex distance(Thielscher and Kammer, 2(02) are required toclarify this important question.

201

Phosphene mapping revealed that stimulation sitesevoking phosphenes bilaterally in the left and rightvisual field depend on the induced current direction.Under the assumption that the generator of an unilat-eral phosphene sits in the contralateral hemispherebilateral phosphenes require a suprathreshold stimu-lation of both occipital lobes. The shift of the borderfrom bilateral to unilateral stimulation can bedirectlyexplained as a consequence of threshold differences.The suprathreshold stimulation of the hemispherestimulated in the preferred current direction stillworks with a coil position shifted over the non-preferred hemisphere. This explanation assumes, inaddition, that the stimulus site under the coil dependson the electric field strength with its maximum atthe midpoint of the coil (Amassian et al., 1992;Ilmoniemi et al., 1999), an assumption that hasrecently been proved for the motor cortex (Thielscherand Kammer, 2(02).

Our data demonstrate that a preference in currentdirection does exist in the visual cortex, comparableto the motor cortex preference. It seems to be ageneral feature of cortical networks to have apreferred current direction since it has been demon-strated in prefrontal cortex, too (Hill et al., 2(00).Studies in the motor system have demonstrated thata change in the current direction changes the responsepattern of the cortical network (Day et al., 1989;Sakai et al., 1997; Di Lazzaro et al., 2(01). It is stillunclear whether different sites of the same popula-tion are stimulated or whether different cellpopulations respond with reversing current direction.Why do we find differences in the cortical responsesin dependence of induced current orientation?Neurons are optimal stimulated with extracellularcurrents oriented in parallel to longitudinal structures,i.e. the axons of the cells (Rushton, 1927). Thestriking dependency on current orientation indicatesan orientation preference ofaxons in the cortex. Inan histological study of the human motor cortex,Marin-Padilla (1970) described a preferential orien-tation of dendritic and axonal fields of large basketcells perpendicular to the main axis of the centralgyrus thus corresponding to the preferential currentdirection in TMS. However, using systematic sections

202

alternatively parallel and perpendicular to the mainaxis of the human precentral gyrus, Meyer (1987)failed to confirm the preferential orientation perpen-dicular but rather found an orientation dominance inparallel to the main gyrus' axis. We are not awareof histological studies of other cortical areas revealinga general orientation preference in relation of thegyral architecture.

In order to explain threshold differences withpolarity reversion of the currents the anatomicalorientation preference has to be accomplished by afunctional anisotropy implementing a diode-likerectifying response behavior. A hypothesis could bederived from the concept of virtual anodessurrounding a stimulating cathode (Rattay, 1987). Ithas been shown that certain geometric arrangementsof electrodes lead to depolarization ofaxons withunidirectional propagation. (Ranck, 1975; Van denHonert and Mortimer, 1979; Ungar et al., 1986). Itremains to be clarified whether the geometry ofelectromagnetically induced currents can mimic sucha behavior.

Acknowledgements

We thank Kuno Kirschfeld, and Hans-GUntherNusseck for support and for many fruitful discus-sions. Michael Erb and Wolfgang Grodd provided uswith anatomical MR scans of the subjects.

References

Amassian, V.E.• Eberle. L., Maccabee, P.I. and Cracco, R.Q.Modelling magnetic coil excitation of human cerebral cortexwith a peripheral nerve immersed in a brain-shaped volumeconductor: the significance of fiber bending in excitation.Electroenceph. Clin. Neurophysiol.• 1992.85: 291-301.

Amassian, V.E.• Maccabee, P.J.• Cracco, R.Q.• Cracco, J.B.•Somasundaram, M.• Rothwell. J.C.• Eberle. L., Henry. K. andRudell. A.P. The polarity of the induced electric field influencesmagnetic coil inhibition of human visual cortex: implicationsfor the site of excitation. Electroenceph. Clin. Neurophysiol .•1994, 93: 21-26.

Barker. A.T.• Garnham, C.W. and Freeston, I.L. Magnetic nervestimulation: the effect of waveform on efficiency. determina-tion of neural membrane time constants and the measurementof stimulator output. Electroenceph. Clin. Neurophysiol., 1991,(Suppl.) 43: 227-237.

Boroojerdi, B.• Meister, I.G.• Foltys, H.• Sparing. R.• Cohen. L.G. and Topper. R. Visual and motor cortex excitability: a tran-scranial magnetic stimulation study. Clin. Neurophysiol.. 2002.113: 1501-1504.

Brasil-Neto, J.P., Cohen. L.G.• Panizza, M.• Nilsson. J., Roth. B.J.and Hallett, M. Optimal focal transcranial magnetic activationof the human motor cortex: effects of coil orientation. shape ofinduced current pulse. and stimulus intensity. J. Clin.Neurophysiol., 1992. 9: 132-136.

Day, B.L.. Dressler. D.• Maertens de Noordhout, A., Marsden. C.D.• Nakashima, K.• Rothwell. J.C. and Thompson, P.D. Electricand magnetic stimulation of human motor cortex: surface EMGand single motor unit responses. J. Physiol. (Lond.). 1989.412:449--473.

Di Lazzaro. V.• Oliviero, A.• Saturno, E.• Pilato. F.• Insola, A.•Mazzone. P., Profice, P., Tonali. P. and Rothwell. lC. Theeffect on corticospinal volleys of reversing the direction ofcurrent induced in the motor cortex by transcranial magneticstimulation. Exp. Brain Res.• 2001, 138: 268-273.

Hill. A.C.. Davey. N.J. and Kennard, C. Current orientationinduced by magnetic stimulation influences a cognitive task.Neuroreport, 2000. 11: 3257-3259.

Ilmoniemi, R.J., Ruohonen, J. and Karhu, J. Transcranial magneticstimulation - a new tool for functional imaging of the brain.Crit. Rev. Biomed. Eng.• 1999. 27: 241-284.

Kammer. T., Beck, S.• Erb, M. and Grodd, W. The influence ofcurrent direction on phosphene thresholds evoked by transcra-nial magnetic stimulation. Clin. Neurophysiol.• 2oo1a. 112:2015-2021.

Kammer. T.• Beck. S.• Thielscher, A.• Laubis-Herrmann, V. andTopka, H. Motor thresholds in humans. A transcranial magneticstimulation study comparing different pulseforrns, current direc-tions and stimulator types. Clin. Neurophysiol.• 2oolb. 112:250--258.

Kozel. FA. Nahas, Z.• deBrux. C.• Molloy. M.• Lorberbaum, J.P.• Bohning. D.• Risch. S.C. and George, M.S. How coil-cortexdistance relates to age, motor threshold. and antidepressantresponse to repetitive transcranial magnetic stimulation.J. Neuropsychiatry Clin. Neurosci.• 2000. 12: 376-384.

Marin-Padilla. M. Prenatal and early postnatal ontogenesis of thehuman motor cortex: a Golgi study. II The basked-pyramidalsystem. Brain Res. 23. 185-191.

McConnell. K. A.• Nahas. Z.• Shastri. A.•Lorberbaum, J. P.• Kozel.F. A.• Bohning, D. E. and George. M. S. (2oo!) The trans-cranial magnetic stimulation motor threshold depends on thedistance from coil to underlying cortex: A replication in healthyadults comparing two methods of assessing the distance tocortex. Bioi. Psychiatry. 1970. 49: 454-459.

Meyer, B.V., Diehl. R.R.• Steinmetz, H., Britton, T.C. andBenecke. R. Magnetic stimuli applied over motor cortex andvisual cortex: Inftuence of coil position and field polarity onmotor responses. phosphenes. and eye movements.Electroenceph. Clin. Neurophysiol.• 1991. (Suppl.) 43: 121-134.

Meyer, G. Forms and Spatial Arrangement of Neurons in thePrimary Motor Cortex of Man. J. Comparative Neurology,1987, 262: 402-428.

Mills, K.R., Boniface, SJ. and Schubert, M. Magnetic brain stim-ulation with a double coil: the importance of coil orientation.Electroenceph. cu« Neurophysiol., 1992, 85, 17-21.

Niehaus, L., Meyer, B.U. and Weyh, T. Influence of pulse config-uration and direction of coil current on excitatory effects ofmagnetic motor cortex and nerve stimulation. Clin.Neurophysiol., 2000, HI: 75-80.

Nithi, K.A. and Mills, K.R. Mapping motor cortex projections tosingle motor units in humans with transcranial magnetic stim-ulation. Muscle Nerve, 2000, 23: 1542-1548.

Ranck, J.B. Which elements are excited in electrical stimulationof mammalian central nervous system: a review. Brain Res.,1975,98: 417-440.

Rattay, F. Ways to approximate current-distance relations for elec-trically stimulated fibers. J. Theor. Bioi., 1987, 125: 339-349.

Rushton, W.A.H. The effect upon the threshold for nervous exci-tation of the length of nerve exposed, and the angle betweencurrent and nerve. J. Physiol. (Lond.), 1927, 63: 357-377.

203

Sakai, K., Ugawa, Y., Terao, Y., Hanajima, R., Furubayashi, T.and Kanazawa, I. Preferential activation of different I wave bytranscranial magnetic stimulation with a figure-of-eight-shapedcoil. Exp. Brain. Res., 1997, 113: 24--32.

Stewart, L.M., Walsh, V. and Rothwell, J.e. Motor and phosphenethresholds: a transeranial magnetic stimulation correlation study.Neuropsychologia, 2001, 39: 415-419.

Thielscher, A. Abschiitzung zum Ort der Nervenstimulation durchMagnetfelder - Ein Beitrag zu den biophysikalischen Grundlagender Transkraniellen Magnetstimulation. Dissertation, AbteilungPsychiatrie ill, Universitat Ulm, 2002.

Thielscher, A. and Kammer, T. Linking Physics with Physiologyin TMS: A sphere field model to determine the cortical stimu-lation site in TMS. Neuroimage, 2002, 17: 1117-1130.

Ungar, lJ., Mortimer, J.T. and Sweeney, J.D. Generation of unidi-rectional propagating action potentials using a monopolarelectrode cuff. Ann. Biomed. Engineer., 1986, 14: 437-450.

Van den Honert, C. and Mortimer, J.T. Generation of unidirec-tionaIly propagated action potentials in a peripheral nerve bybrief stimuli. Science, 1979,206: 1311-1312.