- Email: [email protected]
Regional differences and interindividual variability in sensitivity to vibration in the glabrous skin of the human hand
Brain Research, 301 (1984)65-72 Elsevier
65
BRE 10009
Regional Differences and Interindividual Variability in Sensitivity to Vibration in the Glabrous Skin of the Human Hand J. LC)FVENBERG and R. S. JOHANSSON Department of Physiology, Ume:t University, S-901 87 Ume& (Sweden)
(Accepted October 4th, 1983) Key words: human hand - - detection - - psychophysics - - vibration - - cutaneous mechanoreceptors
Thresholds for sensation of continuous sinusoidal skin displacements were determined in 7 different test points in the glabrous skin area of the right hand of 11 human subjects. While the subjects were tracking the threshold, the frequency of the sine wave was continuously varied between 0.8 Hz and 400 Hz. The obtained threshold-frequency functions showed systematic differences between test points at frequencies below 40-60 Hz. These differences were closely related to density in the skin of the afferent mechanoreceptive units most likely accounting for the decisive afferent signals. At higher frequencies, the interregional variation was less marked whereas there was a pronounced variation between subjects. It was proposed that the detection of the type of stimuli used was based on activity in 3 different mechanoreceptive systems. INTRODUCTION The tactile sensitivity of the human hand is essential both for extracting information about the surrounding world and for control of the motor performance of the hand. Psychophysical studies have revealed regional differences with regard to spatial acuity as well as detection threshold within the hand. The distal half of the distal phalanx stands forth as the most sensitive part in terms of spatial resolution whereas the palm has the lowest capacity in this respect 14,27. Detection threshold to triangular ramp indentations are also higher at the centre of the palm compared to the finger 19. There is a similar proximodistal gradient in density of tactile afferent units innervating the glabrous skin of the human hand 2°. The overall unit density is highest distal to the vortex of the distal phalanx and lowest in the palm. This gradient is mainly accounted for by two of the four types of tactile units in the glabrous skin area: the fast adapting type I (FA I) units, previously denoted rapidly adapting (RA) units (cf. ref. 21), and the slowly adapting type I (SA I) units. The density of these unit types, whose receptive fields are small and well de-
marcated 15, correlates fairly well to the spatial resolution 27. The other two unit types, the fast adapting type II (FA II) units, previously denoted Pacinian (PC) units (cf. ref. 21), and the slowly adapting type II (SA II), have large receptive fields with obscure boundaries and their density is quite uniform within the volar aspect of the hand. Studies of the characteristics of the psychophysical responses to sinusoidal skin displacements of the hand have led to a hypothesis of a duplex mechanism of mechanoreceptionl,5,s,2s,31,32. Correlative studies of psychophysical responses in man and neurophysiological responses in subhuman primates have supported this hypothesis, which postulates that there are two different afferent systems involved in vibrotactile perception, subserved by the two types of fast adapting tactile units, respectively 24-26. Little attention has been paid to regional differences in vibrotactile perception within the hand. The aim of the present study was to examine whether the psychophysicai threshold to sinusoidal skin displacements varies between regions of the glabrous skin of the human hand in the light of unit distribution in the same area. A quick method yielding a continuous
Correspondence: R. S. Johansson, Department of Physiology, University of Ume~, S-901 87 Ume~, Sweden.
0006-8993/84/$03.00 © 1984 Elsevier Science Publishers B.V.
66 threshold function between 0,8 Hz and 400 Hz in 6 min for a single test point was used. METHODS In experiments on 11 healthy human subjects (6 females and 5 males), all in their twenties except two (17 and 36 years), psychophysical thresholds to sinusoidal skin displacements were determined at 7 points in the glabrous skin area of the right hand. None of the subjects had been engaged in heavy manual work. During the experiment the subjects were seated comfortably in a dentist's chair with their right arm extended and supported by a vacuum cast. The dorsal surface of hand was embedded in a Plasticine mould with the fingers moderately flexed. To further minimize the risk of hand movement in relation to the stimulus probe, the proximal phalanges of the ring and middle fingers were fixed with metal clamps, and the index finger was fixed with sticky clay attached to the dorsal surface of the finger. Sinusoidal skin displacements were delivered perpendicularly to the skin surface with a probe which had a flat and circular contact surface with a diameter of 6 mm. A static preindentation of 1 mm was used to obtain stable mechanical coupling between the skin and the probe. The stimulation device was a feedback controlled moving coil stimulator. Its main electrical and mechanical functions have been described elsewhere 33. It was mounted on a heavy metal manipulator, allowing the probe to be oriented perpendicularly to the skin surface at all test points. The input to the stimulator was derived from an analogue multiplier, whose input was a sinewave signal from a function generator and a ramp signal from a ramp generator which thus controlled the amplitude of the output sinewave. Six out of 7 test points were located approximately on a proximo-distal line extending from the centre of the thenar to the very tip of the index finger (Fig. 1). The density of tactile afferent units is known to vary along this line 20. The seventh point was in the centre of the palm. The threshold was determined with a modified version of the von Bekesy threshold tracking method 3,9, i.e., while the frequency of the sinewave continuously changed, the subject exerted control of the amplitude, making it increase or decrease depending on whether he felt the stimulus or not. The subject held
Fig. 1. Location of test points (A-G) on the glabrous skin of the hand. in his free left hand a handle with a pushbutton. When he pressed the button, the ramp generator produced a negative ramp which resulted in a gradual decline of stimulus amplitude and when he released the button the ramp turned positive and the amplitude increased. The subject was instructed to press the button as soon as he felt a movement (upper limen) and to release the button when this sensation disappeared (lower limen). During the test, the frequency of the sine wave was continuously changed from 0.8 Hz to 400 Hz and then back to 0.8 Hz, i.e. it was changed in an ascendingdescending order. This change was exponential with respect to time with a rate of one octave per 20 s. Thus, the total time for a test comprising one ascending and one descending run was 6 min. Before the test proper, the subject was allowed to tune in on the threshold at 0.8 Hz for 20 s. All amplitudes are given in decibels (dB) relative 1 mm peak to peak. The rate of amplitude change increased gradually with decreasing sine wave amplitude, being 2.5 dB/s at - - 2 5 dB and 7.5 dB/s a t - - 6 5 dB. The 7 test points were investigated in a randomized order between subjects, to minimize the effect of training as well as fatigue. With a few exceptions, all points were tested in the same experimental session lasting about 1 h, with short breaks between each test point while the stimulator was repositioned. The experimental data were collected on a X - Y plotter during the experiments. The stimulus amplitude was plotted as a function of the logarithm of the sine wave frequency.
67 at the lowest frequencies, whereas the threshold was minimal at 200-300 Hz (cf. Fig. 3). Moreover, the slope of the function was steepest between 40 and 130 Hz, with a mean fall in amplitude of 12.5 dB per octave (pooled data from all subjects and test points, except the point on the finger tip). For the interval between 4 and 40 Hz the corresponding fall in threshold was 5.8 dB per octave. However, a further analysis indicated large differences between test points as well as between subjects.
-10-
-20®
I
:i
i", =, ,i "'~
'10
.~ -30" t~.
i
40'
b3 -50-
Regional differences
-60-70 ;~
5 1() 2'0 5'0 1(~0 200 Sine wave frequency (c/s)
5(30
Fig. 2. Relation between amplitude and frequency of sinusoidal skin displacements in a threshold tracking task. Test point at the tip of index finger (point G, single subject). Continuous line and segmented line refer to part of test with ascending and descending stimulus frequency, respectively. Stimulus amplitude given in dB relative 1 mm (p-p). Note the logarithmic abscissa.
RESULTS
Fig. 2. shows the original data from one sample test. The solid and the dashed part of the curve refer to the ascending and descending frequency sweep respectively. To facilitate the data analysis, the upper and the lower peaks of the curves were separately joined by straight lines providing an estimate of the upper and the lower limen as a function of stimulus frequency at the ascending and descending sweep, respectively. The average difference between the upper and lower limen was 6 dB at 2.25 Hz, 7 dB at 25 Hz and 8.5 dB at 200 Hz. The threshold was defined as the midpoint between the upper and lower limen. A further data reduction was achieved by estimating the threshold from the graph at 17 different frequencies, spaced approximately equi-distantly on the log-scale. All threshold values referred to below were the mean values of the thresholds of the ascending and the descending run, except for the analysis of differences between the ascending and descending part of the curves (hysteresis). Basically, the threshold frequency functions showed the same general shape as earlier described 2s. Thus, the highest thresholds were obtained
Regional differences in threshold were clearly demonstrated for all 11 subjects. Fig. 3 shows the frequency-threshold functions for 4 of the 7 test points (median values for all subjects). It can be seen that the regional differences were most pronounced at frequencies below 40-60 Hz, where they could roughly be viewed as a proximo-distal gradient with higher thresholds proximally. This is shown by the upper two curves in Fig. 4, which give the thresholds at different points relative to the value at the finger tip at 2 Hz and 25 Hz (median values for all subjects). A comparison between adjacent test points showed particularly pronounced differences between the distal and the proximal part of the distal phalanx, with a median difference between these two test points of 7.5 dB at 2 Hz and 12 dB at 25 Hz. Other pronounced differences were present between the midpalmar point and the two more peripheral points on the
0 a
-10-
" ~- " ",
-20-
-30"
2
-40"
I'-'-
-50" Mid Pa,mar
-60-
F =Proximal part of distal phalanx G = Distal part of distal phalanx
-70-
~
~
~
1~)
2=0
~'~'~.~
5=0 1(~0 2~)
5(~0
Sine wave frequency (c/s) Fig. 3. Psychophysical detection threshold as a function of stimulus frequency at 4 different test points in the glabrous skin area (cf. Fig. 1). Threshold given in dB relative 1 mm (p-p).
68 palm, indicating a notable exception from a strict proximo-distal gradient. Thus, the extreme threshold values were found on the finger tip and in the centre of the palm. The median difference between these two points were 16 dB at 2 Hz and 22 dB at 25 Hz. Statistical tests showed significant threshold differences between the following points (cf. Fig. 1): at 2 Hz, between test point G and all other points and between B and F; and at 25 Hz, between G and all other points, and between B and D, B and E, B and F (P < 0.05 in all cases, simultaneous confidence interval test for paired comparisons34). At frequencies above 40-60 Hz regional differences were less pronounced, but could still be discerned, i.e. the thresholds were clearly higher at the thenar point and in the centre of the palm than on the finger tip (Figs. 3 and 4). However, the maximal difference between any two test points was only 7 dB (B-G) and statistically significant differences at 200 Hz were found only between test points A and G and between B and G (P < 0.05). The shape of the threshold curves for the 3 distal test points on the finger differed distinctly in one respect from the others. To a varying degree they exhibited a plateau in the frequency range 12-60 Hz (cf. Fig. 3). This was most pronounced for the fingertip, where the plateau covered the whole interval. For the two more proximal test points on the finger the flat portion was not as wide and could not be seen
252oE
° 15"
/"
/ ./'.k .
25 Hz#' ~ 2 Hz
"\
~o ~o,
,\ 200 H z l , " ' / / ~ " " ~ ' ~ A
B
I C D Test points
-.
E
F
G
Fig. 4. Median detection threshold at 3 frequencies of stimulation measured at different test points on the glabrous skin area
(cf. Fig. 1). Threshold given in dB relative to threshold on finger tip (test point G).
O10, -20~z -30-4050-6070Sine wave frequency (c/s)
Fig. 5. Interindividual variation of psychophysical detection threshold as a function of sine wave frequency delivered to the finger tip (point G in Fig. 1, 11 subjects). Middle curve gives the median values; hatched area shows the intervals between 25th and 75th percentiles, and top and bottom curves the ranges. Threshold in dB relative 1 mm (p-p).
for all subjects. Even at frequencies below 2-5 Hz, there was occasionally a tendency to a plateau, most often at the finger tip.
Interindividual differences At a given test point, a considerable variation existed between the subjects regarding the thresholds values (Fig. 5). This was particularly prominent for frequencies above 40-60 Hz. The interindividual spread defined as the range was 15 dB, 17 dB and 27 dB at 2 Hz, 25 Hz and 200 Hz, respectively (median range for all test points). A statistical 2-way analysis of variance in threshold was conducted for each of 3 selected frequencies (2, 25 and 200 Hz), with test point (n = 7) and subjects (n = 11) as factors. This analysis showed that the test point was the principal source of variance at low and intermediate frequencies, whereas the principal variance at high frequencies was due to differences between subjects. (The variance explained by the test point factor was 47%, 56% and 11% of the overall variance at 2 Hz, 25 Hz and 200 Hz, respectively. The corresponding variance explained by the subject factor was 27%, 27% and 67%, respectively.) Furthermore, the findings demonstrated that the variability in threshold between subjects could not be described as a simple shift of the wole threshold
69 curves along the ordinate but their shape varied from one subject to the other. On the hypothesis that the sensitivity at one of the 3 selected frequencies was independent of the sensitivity at the other 2, correlation tests between thresholds at 2 Hz, 25 Hz and 200 Hz were performed for each test point (11 subjects). A statistically significant covariation between thresholds at 2 and 25 Hz could only be demonstrated for 3 test points and between 25 and 200 Hz for only 2 test points (P < 0.05, Spearman rank correlation test). At a given frequency, on the other hand, there was a positive correlation between test points in the sense that an individual subject with a low threshold at any given point also showed low thresholds at the other test points compared to other subjects (e.g. P < 0.02 between the finger tip and the centre of the palm, i.e. the two extreme test points, Spearmans rank correlation test).
Quality of sensation After the experiments, some of the subjects were asked to describe the sensations experienced during the tests. They frequently reported 3 different qualities: slow up-down movements of the object in contact with the skin at low stimulus frequencies, a wobbling or fluttery sensation in the skin at higher frequencies, and buzzing and diffuse vibration at the highest frequencies. In separate experiments, carried out mainly on the authors, this was confirmed and the frequencies at which the sensation changed were found to be 4-8 Hz and about 40-60 Hz, respectively. The latter change of sensation occurred at the same range of frequencies as earlier reported26.
Hysteresis Hysteresis was defined as the difference between the two threshold values at any frequency obtained during the limb of ascending and the limb of descending stimulus frequency. The mean value and the range of the hysteresis over the whole frequency range and for all subjects and test points taken together were 1.6 dB and 6 dB respectively. This hysteresis did not appear as a systematic difference between the ascending and descending run in the sense that one was higher. On the other hand, the hysteresis was not completely random; in many runs it appeared to change sign at the segments of the frequency continuum where the quality of sensation shifted.
DISCUSSION On the basis of psychophysical experiments in man and neurophysiological data gathered in subhuman mammals, it has been assumed that two different sets of tactile afferent units innervating the glabrous skin of the human hand are involved in the perception of sinusoidal skin displacements at frequencies above. 5-10 Hz 25,26,31. Tactile units corresponding to the FA I and FA II types in man would mediate the detection of frequencies between ca. 5 and 50 Hz and of frequencies above this range, respectively. This assumption has been supported by recent findings concerning the responses of skin mechanoreceptive units in the human glabrous skin TM. The FA I units are the most sensitive units at midrange frequences, i.e. ca. 8--60 Hz, whereas the corresponding units at higher frequencies are the FA IIs. At frequencies below ca. 5 Hz the detection is most likely based on activity in SA I units since they are most sensitive at those frequencies. A comparison between the psychophysical thresholds obtained at the finger tips in the present study (i.e. the test point showing the lowest thresholds) at 25 Hz, 200 Hz and 2 Hz, and the thresholds of the most sensitive afferent units of the 3 types, respectively, indicate similar values 17As. Thus, the detection within the range of frequencies as studied in the present experiments (0.8-400 Hz) is most likely based on activity in three sets of afferent units. The hysteresis observed in the threshold-frequency functions was probably accounted for by interactions between the proposed 3 different afferent systems during the shift of the system on which the detecting mechanism of the cental nervous system relied. Moreover, the 3 different qualities experienced as well as the the occurrence of two plateaus in the low and mid frequency range respectively, besides the well known U-shape of the function above ca. 60 Hz, would be consistent with a 3 system design. The plateaus between ca. 12 and 60 Hz and below 2-5 Hz observed in the threshold-frequency functions might relate to the corresponding functions for the FA I and SA I units, respectively. The observed proximo-distal gradient in psychophysical vibrotactile sensitivity at low and midrange frequencies corresponds approximately to the gradient in SA I and FA I unit density in the glabrous skin of the median nerve innervation territory, in the
70 sense that the lower the density the higher the threshold. More than half of the threshold gradient was accounted for by the large shift in threshold between the proximal and the distal test points at the terminal phalanx where also the density gradient is steepest 20. Likewise, the highest thresholds were obtained at the palm where the density is lowest. Since there are no reasons to believe that there are substantial variations in thresholds of the afferent units between the different skin regions 19.22, these findings cannot be explained on the basis of a peripheral threshold gradient. In a previous study of the minimal skin indentations necessary to evoke a conscious experience orig'inating from activity in FA I units, the regional pattern was different from that obtained in the present study: the threshold was quite uniform over the whole volar surface of the hand except for the centre of the palm and in the creases where it was higher 19. However, it seems that a perfect agreement between these two studies would not be expected because of the differences in methodology. In the earlier study the subject's task was to detect single triangular ramp indentations without any time constraints, i.e. the detection was made on the basis of a minimal afferent input against a silent background. The results support the hypothesis that the subjects were able to detect a single impulse in a single FA 1 unit in most regions of the hand. Thus the detection was not dependent on the amount of afferent input as long as it exceeded a neural quantum. With the sinusoidal stimulus, on the other hand, as used in the present study, the subject had to perform during a continuous stimulation and within a short time since the amplitude continuously changed. The continuous nature of the stimulus may have accounted for a 'noise' in certain central segments of the somatosensory system which influenced the detection process. According to signal-detection theory, such a 'noise' would give rise to a shift to higher signal levels of the decision criterion as to whether the stimulus is felt or not ~L. Since the density essentially determines the amount of afferent activity, i.e. the signal level, evoked by stimulus amplitudes above the absolute threshold of the population of tactile units 10,23, it seems reasonable that the higher the density, the lower the psychophysical threshold. Likewise, certain masking phenomena 7, perhaps operating on the basis of similar criterion
shifts, may have influenced the psychophysical threshold in the present experiments, since a period of supraliminal intensity of stimulation clearly existed for each cycle of amplitude change. This period was due to the combined effect of the subjects reaction time and the fact that the amplitude continuously changed. If the characteristics (e.g. strength and duration) of possible forward and backward masking effects vary between skin regions, these might also contribute to the interregional variation in psychophysical thresholds. Moreover, at any given test point the reaction time varies with the intensity of the stimulus, and thereby with the amount of afferent activity evoked 25, which in turn is related to the density of afferent units 16,23. On the basis of these considerations it seems, therefore, not surprising that the psychophysical thresholds in this study would decrease from proximal to distal skin regions as the density of FA I and SA I units increases. The fact that the thresholds to sinusoidal stimuli were maximal on the center of the palm compared with those on more peripheral palmar test points agrees with the hypothesis that tactile afferent signals from the central area of the palm are less accurately analysed by the central nervous system 19. Little attention has been focussed on the interindividual variation in vibrotactile sensitivity. However, a comparison between the present results and the few available data from previous studies indicate that the presently found variability was clearly higher than in some studies 25,26, but of the same order of magnitude as reported in other studiesg. Several factors appear to influence the U-shaped part of the threshold-frequency functions at frequencies above ca. 50 Hz, i.e. the limb of the threshold curve accounted for by activity in the FA 11 units. The best known factors are the age of the subject, the size of the contactor ('spatial summation'), the area around the contactor within which the skin can freely move, the size of the static preindentation, the duration of stimulation ('temporal summation') and the skin temperature 6,10,2s-30. Although, except for the temperature, these factors were held relatively constant throughout our experiments, there was still a considerably interindividual variation. In previous studies in our laboratory, the skin temperature of the palm measured in 17 subjects (9 females and 8 males) was found to range between 25 °C and 33 °C tT. According to data published by
71 G r e e n 10 this temperature range can only contribute
explained interindividual variability~, it might be rec-
to ca. 6 dB threshold variation, which is a small quantity compared to the overall interindividual variabili-
o m m e n d e d to use frequencies below ca. 50 Hz in ad-
ty found at any of the test points. Thus, this variability essentially remains unexplained. In contrast the corresponding variability observed at lower frequencies constituted a much smaller fraction since a considerable part of the overall variability at these fre-
dition. Moreover, tests based on low and midrange frequencies may be of particular relevance since the detection at these frquencies most likely will rely on signals in the types of tactile afferent units providing detailed information about spatial aspects of skin stimuli27.
quencies could be explained on the basis of systematic interregional threshold differences. In clinical practice when measuring the skin sensibility using si-
ACKNOWLEDGEMENTS
nusoidal vibrations, high stimulation frequencies are presently used (typically > 100 Hz) 2,4.12,13, i.e. fre-
Financial support by the Swedish Medical Research Council (Project 3548) and the Swedish Work
quencies at which the detection is based on signals in the F A II units. However, as a consequence of the many factors known to influence the detection
E n v i r o n m e n t F u n d is gratefully acknowledged. The informed consent of all subjects was obtained according to the Declaration of Helsinki.
threshold at these frequencies as well as the large un-
REFERENCES 1 Bekesy, G, V., Ober die vibrationsempfindung, Akustische Z., 4 (1939) 316-334. 2 Dellon, A. L., Clinical use of vibratory stimuli to evaluate peripheral nerve injury and compression neuropathy, Plast. reconstruct. Surg., 65 (1980)466-476. 3 Detre, T. P., Feldman, R. G., Rosner, B. and Ferriter, C., Vibration perception in normal and schizophrenic subjects, J. Neuropsychiat., 3 (1962) 145-150. 4 Fagius, J. and Wahren, L.-K., Variability of sensory threshold determinations in clinical use, J. Neurol. Sci., 51 (1981) 11-27. 5 Gescheider, G. A., Evidence in support of the duplex theory of mechanoreception, Sens. Proc., 1 (1976) 68-76. 6 Gescheider, G. A., Capraro, A. J., Frisina, R. D., Hamer, R. D. and Verrillo, R. T., The effects of a surround on vibrotactile thresholds, Sens. Proc., 2 (1978) 99-115. 7 Gescheider, G. A. and Verrillo, R. T., Vibrotactile frequency characteristics as determined by adaptation and masking. In D. R. Kenshalo (Ed.), Sensory Functions of the Skin of Humans, Plenum, New York, 1979, pp. 183-203. 8 Gescheider, G. A., Verrillo, R. T., Capraro, A. J. and Hamar, R. D., Enhancement of vibrotactile sensation magnitude and predictions from the duplex model of mechanoreception, Sens. Proc., 1 (1977) 187-203. 9 Golf, G. D., Rosner, B. S., Detre, T. and Kennard, D., Vibration perception in normal man and medical patients, J. Neurol. Neurosurg. Psychiat., 28 (1965) 503-509. 10 Green, B. G., The effect of skin temperature on vibrotactile sensitivity, Percept. Psychophys., 21 (1977) 243-248. 11 Green, D. M. and Swets, J. A., Signal Detection Theory and Psychophysics, John Wiley and Sons, New York, 1966. 12 Gregersen, G., Vibratory perception threshold and motor conduction velocity in diabetics and non-diabetics, Acta med. scand., 183 (1968) 61-65. 13 Goldberg, J. M. and Lindblom, U,, Standardized method of determining vibratory perception threshold for diagnosis
and screening in neurological investigations, J. Neurol. Neurosurg. Psychiat., 42 (1979) 793-803. 14 Hill, J. W., Limited field of view in reading lettershapes with the fingers. In Geldard (Ed.), Conference of Vibrotactile Communication, The Psychonomic Society, Austin, TX, 1974, pp. 95-105. 15 Johansson, R. S., Tactile sensibility in the human hand: Receptive field characteristics of mechanoreceptive units in the glabrous skin area, J. Physiol. (Lond.), 281 (1978) 101-123. 16 Johansson, R. S., Tactile afferent units with small and well demarcated receptive fields in the glabrous skin area of the human hand. In D. R. Kenshalo (Ed.), Sensory Functions of the Skin of Humans, Plenum, New York, 1979, pp. 129-152. 17 Johansson, R. S., LandstrOm, U. and Lundstr6m, R., Responses of mechanoreceptive afferent units in the glabrous skin of the human hand to sinusoidal skin displacements, Brain Research, 244 (1982) 17-25. 18 Johansson, R. S., Landstr6m, U. and Lundstr6m, R., Sensitivity to edges of mechanoreceptive afferent units in the glabrous skin of the human hand, Brain Research, 244 (1982) 27-32. 19 Johansson, R. S. and Vallbo, /k. B., Detection of tactile stimuli. Thresholds of afferent units related to psychophysical threshold in the human hand, J. Physiol. (Lond.), 297 (1979) 405-422. 20 Johansson, R. S. and Vallbo, ~. B., Tactile sensibility in the human hand: relative and absolute densities of four types of mechanoreceptive units in the glabrous skin area, J. Physiol. (Lond.), 286 (1979) 283-300. 21 Johansson, R. S. and Vallbo, .~. B., Tactile sensory coding in the glabrous skin of the human hand, Trends Neurosci., 6 (1983) 27-32. 22 Johansson, R. S., Vallbo, .~. B. and Westling, G., Thresholds of mechanosensitive afferents in the human hand as measured with von Frey hairs, Brain Research, 184 (1980) 343-351.
72 23 Johnson, K. O., Reconstruction of population response to a vibratory stimulus in quickly adapting mechanoreceptive afferent fiber population innervating glabrous skin of the monkey, J. Neurophysiol., 37 (1974) 48-72. 24 LaMotte, R. H. and Mountcastle, V. B., Capacities of humans and monkeys to discriminate between vibratory stimuli of different frequency and amplitude: a correlation between neural events and psychophysical measurements, J. Neurophysiol., 28 (1975) 966-985. 25 Mountcastle, V. B., LaMotte, R. H. and Carli, G., Detection thresholds for stimuli in humans and monkeys: comparison with threshold events in mechanoreceptive afferent nerve fibres innervating the monkey hand, J. Neurophysiol., 35 (1972) 122-136. 26 Talbot, W. H., Darian-Smith, I., Kornhuber, H. H. and Mountcastle, V. B. The sense of flutter-vibration: comparison of the human capacity with response patterns of mechanoreceptive afferents from the monkey hand, J. Neurophysiol., 31 (1968) 301-334. 27 Vallbo, A. B. and Johansson, R. S., Tactile sensory innervation of the glabrous skin of the human hand. In G. Gordon (Ed.), Active Touch, the Mechanism of Recognition o(
28
29 30 31
32
33
34
Objects by Manipulation, Pergamon Press, Oxford, 1978, pp. 29-54. Verrillo, R. T., Investigation of some parameters of the cutaneous threshold for vibration, J. Acoust. Soc. Amer.. 34 (1962) 1768--1773. Verrillo, R. T., Effect of contractor area on the vibrotactile threshold, J. Acoust. Soc. Amer., 34 35 (1963) 1962-1966. Verrillo, R. T., Temporal summation in vibrotactile sensitivity, J. Acoust. Soc. Amer., 37 (1965) 843--846. Verrillo, R. T., Vibrotactile sensitivity and the frequency response of Pacinian corpuscles, Psychonom. Sci,, 4 (1966) 135-136. Verrillo, R. T., A duplex mechanism of mechanoreception. In D. R. Kenshalo (Ed.), The Skin Senses, Thomas, Springfield, IL, 1968, pp. 139-159. Westling, G.. Johansson, R. S. and Vallbo, ~. B., A method for mechanical stimulation of skin receptors. In Y. Zotterman (Ed.), Sensor>' Functions of the Skin in Primates, Pergamon Press, Oxford, 1976, pp. 151-158. Wonnacott, T. H. and Wonnacott, R. J., Introductory Statistics for Business and Economics, John Wiley and Sons, New York, 1972.
65
BRE 10009
Regional Differences and Interindividual Variability in Sensitivity to Vibration in the Glabrous Skin of the Human Hand J. LC)FVENBERG and R. S. JOHANSSON Department of Physiology, Ume:t University, S-901 87 Ume& (Sweden)
(Accepted October 4th, 1983) Key words: human hand - - detection - - psychophysics - - vibration - - cutaneous mechanoreceptors
Thresholds for sensation of continuous sinusoidal skin displacements were determined in 7 different test points in the glabrous skin area of the right hand of 11 human subjects. While the subjects were tracking the threshold, the frequency of the sine wave was continuously varied between 0.8 Hz and 400 Hz. The obtained threshold-frequency functions showed systematic differences between test points at frequencies below 40-60 Hz. These differences were closely related to density in the skin of the afferent mechanoreceptive units most likely accounting for the decisive afferent signals. At higher frequencies, the interregional variation was less marked whereas there was a pronounced variation between subjects. It was proposed that the detection of the type of stimuli used was based on activity in 3 different mechanoreceptive systems. INTRODUCTION The tactile sensitivity of the human hand is essential both for extracting information about the surrounding world and for control of the motor performance of the hand. Psychophysical studies have revealed regional differences with regard to spatial acuity as well as detection threshold within the hand. The distal half of the distal phalanx stands forth as the most sensitive part in terms of spatial resolution whereas the palm has the lowest capacity in this respect 14,27. Detection threshold to triangular ramp indentations are also higher at the centre of the palm compared to the finger 19. There is a similar proximodistal gradient in density of tactile afferent units innervating the glabrous skin of the human hand 2°. The overall unit density is highest distal to the vortex of the distal phalanx and lowest in the palm. This gradient is mainly accounted for by two of the four types of tactile units in the glabrous skin area: the fast adapting type I (FA I) units, previously denoted rapidly adapting (RA) units (cf. ref. 21), and the slowly adapting type I (SA I) units. The density of these unit types, whose receptive fields are small and well de-
marcated 15, correlates fairly well to the spatial resolution 27. The other two unit types, the fast adapting type II (FA II) units, previously denoted Pacinian (PC) units (cf. ref. 21), and the slowly adapting type II (SA II), have large receptive fields with obscure boundaries and their density is quite uniform within the volar aspect of the hand. Studies of the characteristics of the psychophysical responses to sinusoidal skin displacements of the hand have led to a hypothesis of a duplex mechanism of mechanoreceptionl,5,s,2s,31,32. Correlative studies of psychophysical responses in man and neurophysiological responses in subhuman primates have supported this hypothesis, which postulates that there are two different afferent systems involved in vibrotactile perception, subserved by the two types of fast adapting tactile units, respectively 24-26. Little attention has been paid to regional differences in vibrotactile perception within the hand. The aim of the present study was to examine whether the psychophysicai threshold to sinusoidal skin displacements varies between regions of the glabrous skin of the human hand in the light of unit distribution in the same area. A quick method yielding a continuous
Correspondence: R. S. Johansson, Department of Physiology, University of Ume~, S-901 87 Ume~, Sweden.
0006-8993/84/$03.00 © 1984 Elsevier Science Publishers B.V.
66 threshold function between 0,8 Hz and 400 Hz in 6 min for a single test point was used. METHODS In experiments on 11 healthy human subjects (6 females and 5 males), all in their twenties except two (17 and 36 years), psychophysical thresholds to sinusoidal skin displacements were determined at 7 points in the glabrous skin area of the right hand. None of the subjects had been engaged in heavy manual work. During the experiment the subjects were seated comfortably in a dentist's chair with their right arm extended and supported by a vacuum cast. The dorsal surface of hand was embedded in a Plasticine mould with the fingers moderately flexed. To further minimize the risk of hand movement in relation to the stimulus probe, the proximal phalanges of the ring and middle fingers were fixed with metal clamps, and the index finger was fixed with sticky clay attached to the dorsal surface of the finger. Sinusoidal skin displacements were delivered perpendicularly to the skin surface with a probe which had a flat and circular contact surface with a diameter of 6 mm. A static preindentation of 1 mm was used to obtain stable mechanical coupling between the skin and the probe. The stimulation device was a feedback controlled moving coil stimulator. Its main electrical and mechanical functions have been described elsewhere 33. It was mounted on a heavy metal manipulator, allowing the probe to be oriented perpendicularly to the skin surface at all test points. The input to the stimulator was derived from an analogue multiplier, whose input was a sinewave signal from a function generator and a ramp signal from a ramp generator which thus controlled the amplitude of the output sinewave. Six out of 7 test points were located approximately on a proximo-distal line extending from the centre of the thenar to the very tip of the index finger (Fig. 1). The density of tactile afferent units is known to vary along this line 20. The seventh point was in the centre of the palm. The threshold was determined with a modified version of the von Bekesy threshold tracking method 3,9, i.e., while the frequency of the sinewave continuously changed, the subject exerted control of the amplitude, making it increase or decrease depending on whether he felt the stimulus or not. The subject held
Fig. 1. Location of test points (A-G) on the glabrous skin of the hand. in his free left hand a handle with a pushbutton. When he pressed the button, the ramp generator produced a negative ramp which resulted in a gradual decline of stimulus amplitude and when he released the button the ramp turned positive and the amplitude increased. The subject was instructed to press the button as soon as he felt a movement (upper limen) and to release the button when this sensation disappeared (lower limen). During the test, the frequency of the sine wave was continuously changed from 0.8 Hz to 400 Hz and then back to 0.8 Hz, i.e. it was changed in an ascendingdescending order. This change was exponential with respect to time with a rate of one octave per 20 s. Thus, the total time for a test comprising one ascending and one descending run was 6 min. Before the test proper, the subject was allowed to tune in on the threshold at 0.8 Hz for 20 s. All amplitudes are given in decibels (dB) relative 1 mm peak to peak. The rate of amplitude change increased gradually with decreasing sine wave amplitude, being 2.5 dB/s at - - 2 5 dB and 7.5 dB/s a t - - 6 5 dB. The 7 test points were investigated in a randomized order between subjects, to minimize the effect of training as well as fatigue. With a few exceptions, all points were tested in the same experimental session lasting about 1 h, with short breaks between each test point while the stimulator was repositioned. The experimental data were collected on a X - Y plotter during the experiments. The stimulus amplitude was plotted as a function of the logarithm of the sine wave frequency.
67 at the lowest frequencies, whereas the threshold was minimal at 200-300 Hz (cf. Fig. 3). Moreover, the slope of the function was steepest between 40 and 130 Hz, with a mean fall in amplitude of 12.5 dB per octave (pooled data from all subjects and test points, except the point on the finger tip). For the interval between 4 and 40 Hz the corresponding fall in threshold was 5.8 dB per octave. However, a further analysis indicated large differences between test points as well as between subjects.
-10-
-20®
I
:i
i", =, ,i "'~
'10
.~ -30" t~.
i
40'
b3 -50-
Regional differences
-60-70 ;~
5 1() 2'0 5'0 1(~0 200 Sine wave frequency (c/s)
5(30
Fig. 2. Relation between amplitude and frequency of sinusoidal skin displacements in a threshold tracking task. Test point at the tip of index finger (point G, single subject). Continuous line and segmented line refer to part of test with ascending and descending stimulus frequency, respectively. Stimulus amplitude given in dB relative 1 mm (p-p). Note the logarithmic abscissa.
RESULTS
Fig. 2. shows the original data from one sample test. The solid and the dashed part of the curve refer to the ascending and descending frequency sweep respectively. To facilitate the data analysis, the upper and the lower peaks of the curves were separately joined by straight lines providing an estimate of the upper and the lower limen as a function of stimulus frequency at the ascending and descending sweep, respectively. The average difference between the upper and lower limen was 6 dB at 2.25 Hz, 7 dB at 25 Hz and 8.5 dB at 200 Hz. The threshold was defined as the midpoint between the upper and lower limen. A further data reduction was achieved by estimating the threshold from the graph at 17 different frequencies, spaced approximately equi-distantly on the log-scale. All threshold values referred to below were the mean values of the thresholds of the ascending and the descending run, except for the analysis of differences between the ascending and descending part of the curves (hysteresis). Basically, the threshold frequency functions showed the same general shape as earlier described 2s. Thus, the highest thresholds were obtained
Regional differences in threshold were clearly demonstrated for all 11 subjects. Fig. 3 shows the frequency-threshold functions for 4 of the 7 test points (median values for all subjects). It can be seen that the regional differences were most pronounced at frequencies below 40-60 Hz, where they could roughly be viewed as a proximo-distal gradient with higher thresholds proximally. This is shown by the upper two curves in Fig. 4, which give the thresholds at different points relative to the value at the finger tip at 2 Hz and 25 Hz (median values for all subjects). A comparison between adjacent test points showed particularly pronounced differences between the distal and the proximal part of the distal phalanx, with a median difference between these two test points of 7.5 dB at 2 Hz and 12 dB at 25 Hz. Other pronounced differences were present between the midpalmar point and the two more peripheral points on the
0 a
-10-
" ~- " ",
-20-
-30"
2
-40"
I'-'-
-50" Mid Pa,mar
-60-
F =Proximal part of distal phalanx G = Distal part of distal phalanx
-70-
~
~
~
1~)
2=0
~'~'~.~
5=0 1(~0 2~)
5(~0
Sine wave frequency (c/s) Fig. 3. Psychophysical detection threshold as a function of stimulus frequency at 4 different test points in the glabrous skin area (cf. Fig. 1). Threshold given in dB relative 1 mm (p-p).
68 palm, indicating a notable exception from a strict proximo-distal gradient. Thus, the extreme threshold values were found on the finger tip and in the centre of the palm. The median difference between these two points were 16 dB at 2 Hz and 22 dB at 25 Hz. Statistical tests showed significant threshold differences between the following points (cf. Fig. 1): at 2 Hz, between test point G and all other points and between B and F; and at 25 Hz, between G and all other points, and between B and D, B and E, B and F (P < 0.05 in all cases, simultaneous confidence interval test for paired comparisons34). At frequencies above 40-60 Hz regional differences were less pronounced, but could still be discerned, i.e. the thresholds were clearly higher at the thenar point and in the centre of the palm than on the finger tip (Figs. 3 and 4). However, the maximal difference between any two test points was only 7 dB (B-G) and statistically significant differences at 200 Hz were found only between test points A and G and between B and G (P < 0.05). The shape of the threshold curves for the 3 distal test points on the finger differed distinctly in one respect from the others. To a varying degree they exhibited a plateau in the frequency range 12-60 Hz (cf. Fig. 3). This was most pronounced for the fingertip, where the plateau covered the whole interval. For the two more proximal test points on the finger the flat portion was not as wide and could not be seen
252oE
° 15"
/"
/ ./'.k .
25 Hz#' ~ 2 Hz
"\
~o ~o,
,\ 200 H z l , " ' / / ~ " " ~ ' ~ A
B
I C D Test points
-.
E
F
G
Fig. 4. Median detection threshold at 3 frequencies of stimulation measured at different test points on the glabrous skin area
(cf. Fig. 1). Threshold given in dB relative to threshold on finger tip (test point G).
O10, -20~z -30-4050-6070Sine wave frequency (c/s)
Fig. 5. Interindividual variation of psychophysical detection threshold as a function of sine wave frequency delivered to the finger tip (point G in Fig. 1, 11 subjects). Middle curve gives the median values; hatched area shows the intervals between 25th and 75th percentiles, and top and bottom curves the ranges. Threshold in dB relative 1 mm (p-p).
for all subjects. Even at frequencies below 2-5 Hz, there was occasionally a tendency to a plateau, most often at the finger tip.
Interindividual differences At a given test point, a considerable variation existed between the subjects regarding the thresholds values (Fig. 5). This was particularly prominent for frequencies above 40-60 Hz. The interindividual spread defined as the range was 15 dB, 17 dB and 27 dB at 2 Hz, 25 Hz and 200 Hz, respectively (median range for all test points). A statistical 2-way analysis of variance in threshold was conducted for each of 3 selected frequencies (2, 25 and 200 Hz), with test point (n = 7) and subjects (n = 11) as factors. This analysis showed that the test point was the principal source of variance at low and intermediate frequencies, whereas the principal variance at high frequencies was due to differences between subjects. (The variance explained by the test point factor was 47%, 56% and 11% of the overall variance at 2 Hz, 25 Hz and 200 Hz, respectively. The corresponding variance explained by the subject factor was 27%, 27% and 67%, respectively.) Furthermore, the findings demonstrated that the variability in threshold between subjects could not be described as a simple shift of the wole threshold
69 curves along the ordinate but their shape varied from one subject to the other. On the hypothesis that the sensitivity at one of the 3 selected frequencies was independent of the sensitivity at the other 2, correlation tests between thresholds at 2 Hz, 25 Hz and 200 Hz were performed for each test point (11 subjects). A statistically significant covariation between thresholds at 2 and 25 Hz could only be demonstrated for 3 test points and between 25 and 200 Hz for only 2 test points (P < 0.05, Spearman rank correlation test). At a given frequency, on the other hand, there was a positive correlation between test points in the sense that an individual subject with a low threshold at any given point also showed low thresholds at the other test points compared to other subjects (e.g. P < 0.02 between the finger tip and the centre of the palm, i.e. the two extreme test points, Spearmans rank correlation test).
Quality of sensation After the experiments, some of the subjects were asked to describe the sensations experienced during the tests. They frequently reported 3 different qualities: slow up-down movements of the object in contact with the skin at low stimulus frequencies, a wobbling or fluttery sensation in the skin at higher frequencies, and buzzing and diffuse vibration at the highest frequencies. In separate experiments, carried out mainly on the authors, this was confirmed and the frequencies at which the sensation changed were found to be 4-8 Hz and about 40-60 Hz, respectively. The latter change of sensation occurred at the same range of frequencies as earlier reported26.
Hysteresis Hysteresis was defined as the difference between the two threshold values at any frequency obtained during the limb of ascending and the limb of descending stimulus frequency. The mean value and the range of the hysteresis over the whole frequency range and for all subjects and test points taken together were 1.6 dB and 6 dB respectively. This hysteresis did not appear as a systematic difference between the ascending and descending run in the sense that one was higher. On the other hand, the hysteresis was not completely random; in many runs it appeared to change sign at the segments of the frequency continuum where the quality of sensation shifted.
DISCUSSION On the basis of psychophysical experiments in man and neurophysiological data gathered in subhuman mammals, it has been assumed that two different sets of tactile afferent units innervating the glabrous skin of the human hand are involved in the perception of sinusoidal skin displacements at frequencies above. 5-10 Hz 25,26,31. Tactile units corresponding to the FA I and FA II types in man would mediate the detection of frequencies between ca. 5 and 50 Hz and of frequencies above this range, respectively. This assumption has been supported by recent findings concerning the responses of skin mechanoreceptive units in the human glabrous skin TM. The FA I units are the most sensitive units at midrange frequences, i.e. ca. 8--60 Hz, whereas the corresponding units at higher frequencies are the FA IIs. At frequencies below ca. 5 Hz the detection is most likely based on activity in SA I units since they are most sensitive at those frequencies. A comparison between the psychophysical thresholds obtained at the finger tips in the present study (i.e. the test point showing the lowest thresholds) at 25 Hz, 200 Hz and 2 Hz, and the thresholds of the most sensitive afferent units of the 3 types, respectively, indicate similar values 17As. Thus, the detection within the range of frequencies as studied in the present experiments (0.8-400 Hz) is most likely based on activity in three sets of afferent units. The hysteresis observed in the threshold-frequency functions was probably accounted for by interactions between the proposed 3 different afferent systems during the shift of the system on which the detecting mechanism of the cental nervous system relied. Moreover, the 3 different qualities experienced as well as the the occurrence of two plateaus in the low and mid frequency range respectively, besides the well known U-shape of the function above ca. 60 Hz, would be consistent with a 3 system design. The plateaus between ca. 12 and 60 Hz and below 2-5 Hz observed in the threshold-frequency functions might relate to the corresponding functions for the FA I and SA I units, respectively. The observed proximo-distal gradient in psychophysical vibrotactile sensitivity at low and midrange frequencies corresponds approximately to the gradient in SA I and FA I unit density in the glabrous skin of the median nerve innervation territory, in the
70 sense that the lower the density the higher the threshold. More than half of the threshold gradient was accounted for by the large shift in threshold between the proximal and the distal test points at the terminal phalanx where also the density gradient is steepest 20. Likewise, the highest thresholds were obtained at the palm where the density is lowest. Since there are no reasons to believe that there are substantial variations in thresholds of the afferent units between the different skin regions 19.22, these findings cannot be explained on the basis of a peripheral threshold gradient. In a previous study of the minimal skin indentations necessary to evoke a conscious experience orig'inating from activity in FA I units, the regional pattern was different from that obtained in the present study: the threshold was quite uniform over the whole volar surface of the hand except for the centre of the palm and in the creases where it was higher 19. However, it seems that a perfect agreement between these two studies would not be expected because of the differences in methodology. In the earlier study the subject's task was to detect single triangular ramp indentations without any time constraints, i.e. the detection was made on the basis of a minimal afferent input against a silent background. The results support the hypothesis that the subjects were able to detect a single impulse in a single FA 1 unit in most regions of the hand. Thus the detection was not dependent on the amount of afferent input as long as it exceeded a neural quantum. With the sinusoidal stimulus, on the other hand, as used in the present study, the subject had to perform during a continuous stimulation and within a short time since the amplitude continuously changed. The continuous nature of the stimulus may have accounted for a 'noise' in certain central segments of the somatosensory system which influenced the detection process. According to signal-detection theory, such a 'noise' would give rise to a shift to higher signal levels of the decision criterion as to whether the stimulus is felt or not ~L. Since the density essentially determines the amount of afferent activity, i.e. the signal level, evoked by stimulus amplitudes above the absolute threshold of the population of tactile units 10,23, it seems reasonable that the higher the density, the lower the psychophysical threshold. Likewise, certain masking phenomena 7, perhaps operating on the basis of similar criterion
shifts, may have influenced the psychophysical threshold in the present experiments, since a period of supraliminal intensity of stimulation clearly existed for each cycle of amplitude change. This period was due to the combined effect of the subjects reaction time and the fact that the amplitude continuously changed. If the characteristics (e.g. strength and duration) of possible forward and backward masking effects vary between skin regions, these might also contribute to the interregional variation in psychophysical thresholds. Moreover, at any given test point the reaction time varies with the intensity of the stimulus, and thereby with the amount of afferent activity evoked 25, which in turn is related to the density of afferent units 16,23. On the basis of these considerations it seems, therefore, not surprising that the psychophysical thresholds in this study would decrease from proximal to distal skin regions as the density of FA I and SA I units increases. The fact that the thresholds to sinusoidal stimuli were maximal on the center of the palm compared with those on more peripheral palmar test points agrees with the hypothesis that tactile afferent signals from the central area of the palm are less accurately analysed by the central nervous system 19. Little attention has been focussed on the interindividual variation in vibrotactile sensitivity. However, a comparison between the present results and the few available data from previous studies indicate that the presently found variability was clearly higher than in some studies 25,26, but of the same order of magnitude as reported in other studiesg. Several factors appear to influence the U-shaped part of the threshold-frequency functions at frequencies above ca. 50 Hz, i.e. the limb of the threshold curve accounted for by activity in the FA 11 units. The best known factors are the age of the subject, the size of the contactor ('spatial summation'), the area around the contactor within which the skin can freely move, the size of the static preindentation, the duration of stimulation ('temporal summation') and the skin temperature 6,10,2s-30. Although, except for the temperature, these factors were held relatively constant throughout our experiments, there was still a considerably interindividual variation. In previous studies in our laboratory, the skin temperature of the palm measured in 17 subjects (9 females and 8 males) was found to range between 25 °C and 33 °C tT. According to data published by
71 G r e e n 10 this temperature range can only contribute
explained interindividual variability~, it might be rec-
to ca. 6 dB threshold variation, which is a small quantity compared to the overall interindividual variabili-
o m m e n d e d to use frequencies below ca. 50 Hz in ad-
ty found at any of the test points. Thus, this variability essentially remains unexplained. In contrast the corresponding variability observed at lower frequencies constituted a much smaller fraction since a considerable part of the overall variability at these fre-
dition. Moreover, tests based on low and midrange frequencies may be of particular relevance since the detection at these frquencies most likely will rely on signals in the types of tactile afferent units providing detailed information about spatial aspects of skin stimuli27.
quencies could be explained on the basis of systematic interregional threshold differences. In clinical practice when measuring the skin sensibility using si-
ACKNOWLEDGEMENTS
nusoidal vibrations, high stimulation frequencies are presently used (typically > 100 Hz) 2,4.12,13, i.e. fre-
Financial support by the Swedish Medical Research Council (Project 3548) and the Swedish Work
quencies at which the detection is based on signals in the F A II units. However, as a consequence of the many factors known to influence the detection
E n v i r o n m e n t F u n d is gratefully acknowledged. The informed consent of all subjects was obtained according to the Declaration of Helsinki.
threshold at these frequencies as well as the large un-
REFERENCES 1 Bekesy, G, V., Ober die vibrationsempfindung, Akustische Z., 4 (1939) 316-334. 2 Dellon, A. L., Clinical use of vibratory stimuli to evaluate peripheral nerve injury and compression neuropathy, Plast. reconstruct. Surg., 65 (1980)466-476. 3 Detre, T. P., Feldman, R. G., Rosner, B. and Ferriter, C., Vibration perception in normal and schizophrenic subjects, J. Neuropsychiat., 3 (1962) 145-150. 4 Fagius, J. and Wahren, L.-K., Variability of sensory threshold determinations in clinical use, J. Neurol. Sci., 51 (1981) 11-27. 5 Gescheider, G. A., Evidence in support of the duplex theory of mechanoreception, Sens. Proc., 1 (1976) 68-76. 6 Gescheider, G. A., Capraro, A. J., Frisina, R. D., Hamer, R. D. and Verrillo, R. T., The effects of a surround on vibrotactile thresholds, Sens. Proc., 2 (1978) 99-115. 7 Gescheider, G. A. and Verrillo, R. T., Vibrotactile frequency characteristics as determined by adaptation and masking. In D. R. Kenshalo (Ed.), Sensory Functions of the Skin of Humans, Plenum, New York, 1979, pp. 183-203. 8 Gescheider, G. A., Verrillo, R. T., Capraro, A. J. and Hamar, R. D., Enhancement of vibrotactile sensation magnitude and predictions from the duplex model of mechanoreception, Sens. Proc., 1 (1977) 187-203. 9 Golf, G. D., Rosner, B. S., Detre, T. and Kennard, D., Vibration perception in normal man and medical patients, J. Neurol. Neurosurg. Psychiat., 28 (1965) 503-509. 10 Green, B. G., The effect of skin temperature on vibrotactile sensitivity, Percept. Psychophys., 21 (1977) 243-248. 11 Green, D. M. and Swets, J. A., Signal Detection Theory and Psychophysics, John Wiley and Sons, New York, 1966. 12 Gregersen, G., Vibratory perception threshold and motor conduction velocity in diabetics and non-diabetics, Acta med. scand., 183 (1968) 61-65. 13 Goldberg, J. M. and Lindblom, U,, Standardized method of determining vibratory perception threshold for diagnosis
and screening in neurological investigations, J. Neurol. Neurosurg. Psychiat., 42 (1979) 793-803. 14 Hill, J. W., Limited field of view in reading lettershapes with the fingers. In Geldard (Ed.), Conference of Vibrotactile Communication, The Psychonomic Society, Austin, TX, 1974, pp. 95-105. 15 Johansson, R. S., Tactile sensibility in the human hand: Receptive field characteristics of mechanoreceptive units in the glabrous skin area, J. Physiol. (Lond.), 281 (1978) 101-123. 16 Johansson, R. S., Tactile afferent units with small and well demarcated receptive fields in the glabrous skin area of the human hand. In D. R. Kenshalo (Ed.), Sensory Functions of the Skin of Humans, Plenum, New York, 1979, pp. 129-152. 17 Johansson, R. S., LandstrOm, U. and Lundstr6m, R., Responses of mechanoreceptive afferent units in the glabrous skin of the human hand to sinusoidal skin displacements, Brain Research, 244 (1982) 17-25. 18 Johansson, R. S., Landstr6m, U. and Lundstr6m, R., Sensitivity to edges of mechanoreceptive afferent units in the glabrous skin of the human hand, Brain Research, 244 (1982) 27-32. 19 Johansson, R. S. and Vallbo, /k. B., Detection of tactile stimuli. Thresholds of afferent units related to psychophysical threshold in the human hand, J. Physiol. (Lond.), 297 (1979) 405-422. 20 Johansson, R. S. and Vallbo, ~. B., Tactile sensibility in the human hand: relative and absolute densities of four types of mechanoreceptive units in the glabrous skin area, J. Physiol. (Lond.), 286 (1979) 283-300. 21 Johansson, R. S. and Vallbo, .~. B., Tactile sensory coding in the glabrous skin of the human hand, Trends Neurosci., 6 (1983) 27-32. 22 Johansson, R. S., Vallbo, .~. B. and Westling, G., Thresholds of mechanosensitive afferents in the human hand as measured with von Frey hairs, Brain Research, 184 (1980) 343-351.
72 23 Johnson, K. O., Reconstruction of population response to a vibratory stimulus in quickly adapting mechanoreceptive afferent fiber population innervating glabrous skin of the monkey, J. Neurophysiol., 37 (1974) 48-72. 24 LaMotte, R. H. and Mountcastle, V. B., Capacities of humans and monkeys to discriminate between vibratory stimuli of different frequency and amplitude: a correlation between neural events and psychophysical measurements, J. Neurophysiol., 28 (1975) 966-985. 25 Mountcastle, V. B., LaMotte, R. H. and Carli, G., Detection thresholds for stimuli in humans and monkeys: comparison with threshold events in mechanoreceptive afferent nerve fibres innervating the monkey hand, J. Neurophysiol., 35 (1972) 122-136. 26 Talbot, W. H., Darian-Smith, I., Kornhuber, H. H. and Mountcastle, V. B. The sense of flutter-vibration: comparison of the human capacity with response patterns of mechanoreceptive afferents from the monkey hand, J. Neurophysiol., 31 (1968) 301-334. 27 Vallbo, A. B. and Johansson, R. S., Tactile sensory innervation of the glabrous skin of the human hand. In G. Gordon (Ed.), Active Touch, the Mechanism of Recognition o(
28
29 30 31
32
33
34
Objects by Manipulation, Pergamon Press, Oxford, 1978, pp. 29-54. Verrillo, R. T., Investigation of some parameters of the cutaneous threshold for vibration, J. Acoust. Soc. Amer.. 34 (1962) 1768--1773. Verrillo, R. T., Effect of contractor area on the vibrotactile threshold, J. Acoust. Soc. Amer., 34 35 (1963) 1962-1966. Verrillo, R. T., Temporal summation in vibrotactile sensitivity, J. Acoust. Soc. Amer., 37 (1965) 843--846. Verrillo, R. T., Vibrotactile sensitivity and the frequency response of Pacinian corpuscles, Psychonom. Sci,, 4 (1966) 135-136. Verrillo, R. T., A duplex mechanism of mechanoreception. In D. R. Kenshalo (Ed.), The Skin Senses, Thomas, Springfield, IL, 1968, pp. 139-159. Westling, G.. Johansson, R. S. and Vallbo, ~. B., A method for mechanical stimulation of skin receptors. In Y. Zotterman (Ed.), Sensor>' Functions of the Skin in Primates, Pergamon Press, Oxford, 1976, pp. 151-158. Wonnacott, T. H. and Wonnacott, R. J., Introductory Statistics for Business and Economics, John Wiley and Sons, New York, 1972.