Exp. Brain Res. 22, 331--334 (1975) �9 by Splqnger-Verlag 1975

Somatosensory Evoked Potentials and Magnitude of Perception

D. JOHNSON, R. JORGENS, (]-. KO~qGEHL and H. It . KOI~NHU~E~ University of Ulm, Ulm (FRG)

Received December 15, 1974

Summary. With step indentations of the index finger tip in randomized order, via a mechanostimu]ator, the tactile receptors of human skin were adequately stimulated. Recording the EEG over the contralateral and ipsilateral cortex, the evoked potentials and their 95% confidence limits were analysed. Simultaneously the psychophysical magnitude estimations were analysed.

1. The perceptual estimations were linearly related to step amplitude. 2. The early components of the E.P. show no obvious correlation to stimulus

amplitude. 3. The later components (with peak latencies of 120 msec or more) show a

monotone, non-linear rising function with respect to stimulus amplitude. 4. The early waves of the evoked potentials up to about 120 msec are well

localized over the contralateral postcentral hand area while the late components resemble the alpha rhy thm in wave length and distribution over both hemispheres.

The possible role of alpha-synchronisation in the later components is discussed.

Key words: Somatosensory - - Evoked potentials - - Psychophysics - - Merkel discs - - Man

During the last decade, the late components of cortical evoked potentials (E.P.) with peak latencies of 120 ms or more have been considered to reflect the neural processes underlying perception. Power function relationships have been found between response and stimulus intensity with the same exponent for the late components of the E.P. and the psychophysical estimations (Keidel and Spreng, 1965; FranzSn and Offenloch, 1969).

We have investigated the somatosensory system, in which the physiological bases are well understood from the peripheral receptor up to the cortical neurons. Based on their different frequency response characteristics, we can nearly selecti- vely stimulate three receptor sets in the glabrous skin of the human finger tip. The only receptor type with large myelinated fibers able to discriminate steps of different indentation magnitudes is the slowly adapting Merkcl cell (Mountcastle, 1968; Kornhuber, 1972). These receptors are known to have a strictly linear stimulus-response relation ()/fountcastle et al., 1966). The cortical neurons in the hand area of Rhesus postcentral gyrus with Merkel afferents also show the same linear response (Mountcastle, (1968). We have tried to establish the quanti tat ive relationship between the evoked potentials and the magnitude of perception for the step indentation stimulus.

332 D. Johnson et al.

Methods

With 9 normal, awake human subjects, a step indentation of the finger tip was used as the adequate mechanical stimulus for the S.A. Merkel receptors. Randomized stimulus inten- sities were applied via a mechanostimulator (Burchard et al., 1967) which drove a plexiglas probe (1 mm contact diameter). Four different magnitude categories were given per experi- ment, with 128 randomized repetitions. The indentation was automatically controlled and the course of the applied force was simultaneously recorded. The start position in relation to the skin was also automatically reset prior to the stimuli. Auditory artefacts were masked in all cases. Besides the Merkel discs, the other receptors will also respond to the stimulus but only with one or two action potentials and thus could only play a minor role in perception of magni- tude (Talbot et al., 1968). The stimulus was maintained throughout the response to avoid overlap of "off-effects". The rise time of the step with the finger on the probe was 5 msec for the smallest (10O/2m) and 12 msee for the largest (1.000/zm) stimulus.

The cerebral evoked potentials were recorded monopolarly over the hand area of the post- central gyrus of both hemispheres (reference electrode at joined ears) from the scalp with standard EEG electrodes and off-line computer averaging. Via reverse tape analysis and after elimination of artefacts the E.P.'s were averaged and the 95O/o confidence limits were calculat- ed for each recording (Decus program 12--98). The baseline prior to the stimulus, as an amplitude reference level, was rejected and only peak-to-peak measurements were used. Simultaneously, the psychophysical estimations from the same experiments were analysed. (Method of absolute Judgement).

Results

As can be readily seen from Fig. 1, the psychophysical st imulus-response

relat ion is l inear under these exper imenta l conditions. I n contrast to this and to the

peripheral receptors, to the cortical neurons and to perception, the evoked potent ia l

components, show no linear relat ion between their ampl i tudes and st imulus magni-

tude (Fig. 2). The earliest waves (N25-P59) show no relat ionship at all to s t imulus

magni tude. The differences in waveform of these early components , as compared

to those of other authors, may be adequa te ly explained by the different st imuli

"ABSOLUTE" JUDGEMENT

f 9 SUBJ.

RI. iNDEX FINGER TIP

100 350 600 850 ,,u SKIN INDENTATION

Fig. 1. Psychophysical perception curve. Estimated magnitude of stimuli, versus step skin indentation, normalized average of 9 subjects

Somatosensory Evoked Potentials and Perception 333

A

N158 i [IO.uV

50 msec

-L:2sE P319

P 59

N=128

STEP INDENTATION I =950 Mm

PEAK

pV TO PEAK

AMPLITUDE P 59 - N 158

15

10 N lS8 - P 2 0 6

N 245- P 319

5 R 206 - N 245

~ N 25-P 59

200 400 600 800 10()0 # m L

DEPTH OF FINGER TIP iNDENTATION

Fig. 2. (A) Averaged evoked potential following 128 step indentations of the right index finger tip from one typical subject. Monopolar recording, contralateral, posteentral hand area versus joined ears. Surface negativity upwards. Peak latencies of the components are given in msec SE Standard error. The time course of the stimulus is shown below. (B) Amplitude curves of the different components of the somatosensory evoked potentials for the same subject. The vertical bars give +2 x standard error, calculated from the 128 responses for each stimulus magnitude

used. The la te r waves, however, do show a sys temat ic re la t ionship to the s t imulus a t the 95% significance level as seen in Fig. 2, bu t no t a l inear one. Al l s t imulus response curves h a d a p p r o x i m a t e l y the same form and var iance. P lo t t ing s t imulus force as abscissa, ins tead of inden ta t ion , produces no change in form of the curves. The peak latencies of all waves decrease wi th increasing s t imulus in tens i ty , bu t on ly to a m i n i m u m b y a b o u t 300 # m inden ta t ion . The la te components , wi th onset g rea te r t h a n 120 reset , are found b i l a te ra l ly over the whole scalp while the ear ly componen t s are s t r ic t ly cont ra la te ra l . The la te components have ampl i tudes in- creasing f rom preccn t ra l t owards occipi tal .

Discussion

Al though the magn i tude percep t ion shows the same l inear re la t ion to s t imulus in t ens i ty as do the cort ical neurons, the E. P . - eomponen t s do no t reflect this s imple re la t ion. Ac tua l ly , a l inear re la t ion of the evoked po ten t i a l ampl i tude , recorded

334 D. Johnson et al.

from the scalp, to single uni t firing f requency in the cor tex should no t be expec ted because of the neuronal and phys ica l complexi ty . The la te components in the somatosensory E . P . have a typ ica l f requency of a b o u t 10 Hz ( approx ima te ly a lpha frequency) and are widely d i s t r ibu ted over the cor tex wi th increasing ampl i tudes towards the occipi ta l lobe as do those of the visual E . P . whereas the a u d i t o r y E . P . has a f requency of abou t 5 Hz, t h a t is the first subharmonie of the a lpha r h y t h m . F r o m clinical cases, we also know t h a t the occipi ta l and f ronta l lobes as well as the whole ips i la tera l hemispheres are no t requi red for t ac t i l e per- ception. Due to the i r f requency and d i s t r ibu t ion we suspect these la te components m a y represent a non-specific synchronisa t ion of tha lamo-cor t i ca l a lpha generators and are p r o b a b l y not d i rec t ly re la ted to sensat ion. A possible significance for percept ion of some ampl i tude components h idden in the la te waves is, however, no t con t rad ic ted b y this hypothesis .

Acknowledgement: The authors wish to express their appreciation for invaluable assistance and advice to Mr. W. Becker.

References

Burchard, D., Kapp, H., Kornhuber, H. H. : Ein Kraft- und Weg-kontrolliertes mechanisches Reizgerat fiir Untersuchungen der somatischen Sensiblititi~t. Pfliigers Arch. ges. Physiol. 297, 99 (1967)

Franz~n, 0., Offenloch, K. : Evoked response correlates of psychophysical magnitude estimates for tactile stimulation in man. Exp. Brain Res. 8, 1--18 (1969)

Keidel, W.B., Spreng, M.: Neurophysiological evidence for the Stevens power function in man. J. acoust. Soc. Amer. 38, 191--195 (1965)

Kornhuber, H.H. : Tastsinn und Lagesinn. In: Gauer, Kramer, Jung (Eds.): Physiologie des Menschen, Vol. 11, S. 52--112. Miinchen: Urban und Schwarzenberg 1972

Mountcastle, V.B. (Ed.): Medical Physiology, 12th Ed., Vol. II. St. Louis: C.V. 1VIosby 1968 Mountcastle, V. B., Talbot, W. It., Kornhuber, H. H. : The neural transformation of mechanical

stimuli delivered to the monkey's hand. In: De Reuck and Knight (Eds.) : CIBA Foundation Symposium on touch, he~t and pain, p. 325--345. London: Churchill 1966

Talbot, W.H., Darian-Smith, I., Kornhuber, H.H., Mountcastle, V. B. : The sense of flutter- vibration. J. Neurophysiol. 31, 301--334 (1968)

D. Johnson University of Ulm Sect. Neurophysiology D - 79 Ulm Federal Republic of Germany