Here, we present protocols to determine vibration detection thresholds and tactile acuity using psychophysical methods in man.
Tests that allow the precise determination of psychophysical thresholds for vibration and grating orientation provide valuable information about mechanosensory function that are relevant for clinical diagnosis as well as for basic research. Here, we describe two psychophysical tests designed to determine the vibration detection threshold (automated system) and tactile spatial acuity (handheld device). Both procedures implement a two-interval forced-choice and a transformed-rule up and down experimental paradigm. These tests have been used to obtain mechanosensory profiles for individuals from distinct human cohorts such as twins or people with sensorineural deafness.
Specialized mechanosensory receptors in the skin mediate the perception of vibration and grating orientation. Each mechanosensory receptor type is tuned to detect distinct features of tactile stimuli1,2. This property provides the psychophysical basis for a differentiated assessment of mechanosensory function by using tests that deliver simple sinusoidal oscillations (vibration) or fine gratings. It is known that psychophysical thresholds for vibration perception are lower for high than low frequency vibration3,4. Two types of rapidly adapting mechanoreceptors associated with the Meissner and Pacinian corpuscles primarily detect low (10-40 Hz) and high (100-200 Hz) frequency vibration stimuli, respectively5. It is thought that the psychophysical thresholds for vibration perception rely to a considerable extent on the activation of these two sets of mechanoreceptors at their best frequencies6,7. Tactile spatial acuity is tested by the grating orientation task to determine the finest grating whose orientation can be discriminated by a subject7–9. Merkel cell-neurite complex afferents are essential for detecting grating orientation8,10. Interestingly, there has been considerable recent progress in our understanding of the molecular basis of how mechanoreceptors detect extremely small tactile stimuli11. Mechanosensitive ion channels like Piezo2 and modulators like STOML3 have been directly implicated in the detection of fine tactile stimuli in touch receptors12–15. There are already human patients identified with function altering mutations in the Piezo2 gene and it will be important to test whether such patients have touch deficits16.
The determination of the vibrotactile threshold relies on the delivery of a vibrating stimulus to the skin. Vibration detection thresholds vary with the vibration frequency. Vibration frequency to detection threshold curves were described by von Bekesey, Verillo, and Bolanowski, among many others, from the 1930s up to the 1990s5,17. The devices in the early days were based on shakers and power amplifiers and several devices have been commercialized and there are a lot of variation as to choice of stimulator (testing frequency), the diameter of the probe in contact with skin, the use of a surround to limit the stimulation to one area of the skin, and the testing protocol that determines the threshold (see the following references for more detailed insight on the device and testing features)18–20. Most devices usually test one site; however, there are new devices that use vertical displacement stimulators equipped with 2 probes to deliver vibratory stimuli with separation ranges that can be varied21. Also detection thresholds are measured based on vibration frequency or intensity discrimination with continuous stimulation; or with intermittent vibration stimuli with or without a masking stimulus. Therefore, we recommend that the reader be aware of the plethora of developments in this field.
Here, we describe the components of the device and a step-by-step guide on how to conduct psychophysical tests to estimate vibration detection threshold in human subjects. We then show how to assess tactile acuity manually by using the tactile acuity cube. We use a two-interval, forced-choice paradigm: the stimulus is always presented during one of two intervals and the subject has to indicate which interval has the vibration stimulus. We employ a transformed-rule up and down as the adaptive method that executes a threshold search by changing the stimulus intensity based on the performance during the test. Both psychophysical protocols can be used by investigators as a screening tool for evaluation of alterations in touch sensitivity.
The testing protocol was approved by the Charité-Universitätsmedizin Ethics Committee.
1. Vibration Detection Threshold (VDT)
2. Tactile Spatial Acuity Test
The piezoelectric actuator provides the vibration stimulus to the subject. The vibration stimulus has a total duration of 1.8 sec and is presented only once during a trial during the first or second interval (Figure 2A). The rise and fall time at the onset and offset of the stimulus is determined by the functions (1-e-bt)∙Amplitude∙sine(frequency), and (e-bt)∙Amplitude∙sine(frequency), respectively, where b is set at 9.1. The rise and fall time at onset and offset are 500 and 600 msec, respectively, and are independent of the testing frequency and amplitude. The duration of the stimulus between the onset and offset phases is 700 msec. The gradual rise and fall ensures a smooth stimulus delivery.
Vibration detection thresholds are dependent on stimulation frequency because they are mediated by different sensory receptors. According to the human psychophysical tuning curve, thresholds lie between ~20 nm to ~45 µm5. Therefore, a set of 35 levels of stimulation (range from 18 nm to 45 µm) of the vibration waveform is constructed (Figure 2B) whose amplitude values are arranged logarithmically (base 10; stimulusn+1 = 100.1∙stimulusn). This range of amplitudes is designed to allow to test at frequencies ranging from 1-250 Hz. The starting stimulus amplitude is usually set above the average vibration detection threshold for a particular vibration frequency test. Previous observations on frequency tuning curves obtained from both psychophysical studies measured average detection threshold around 300 nm for high frequency (>100 Hz) stimuli, and ~3 µm for lower frequencies (<40 Hz)5,23,24.
The relationship of the driving input voltage to the output displacement of the piezoelectric actuator measured by the strain gauge sensor (SGS) at 10 and 125 Hz are illustrated in Figures 3A and 3B. The relationship is linear (correlation coefficient, r2 = 0.9992) for the set of vibration intensities used (Figure 3C). There was a nearly identical relationship between input voltage and output displacement at sinusoidal waveforms of 20 nm amplitude (Figure 3D).
In Figure 4, a typical testing session where the vibration detection threshold at the right little finger is determined at 125 Hz. The threshold search starts with a stimulus that is above threshold of amplitude level 23 (676 nm displacement). The experimental session consists of multiple series of single trials. A series consists of up to 9 single trials applied at one vibration stimulus level. A decrease in stimulus intensity (level) in general requires at least 6 sequential correct responses (Figure 4A). A change in the direction of stimulus intensity marks a reversal point in trial series track and requires at least two incorrect responses. The magnitude of change in stimulus intensity depends on the reversal number. To rapidly search for the threshold, the stimulus intensity is changed in steps of 4 levels when the number of reversals is less than 3, followed by 2 levels at the 3rd reversal (upward direction). Otherwise, the stimulus intensity changes in 1-level steps to finely determine the threshold (Figure 4B). The subject's threshold is calculated by converting the median stimulus intensity of the last 6 reversal points; in this case 401 nm. The transformed-rule up and down procedure converges to the threshold value at which 75% of the responses are correct. This was calculated based on the response sequences leading to a down rule using the following formula: p6 + 6p6.(1-p)+6p6.(1-p)2, where p is the probability of a correct response.
In the same subject, the tactile cube experiment was done (Figure 5). Each trial consists of 2 stimulations. A two-down and one-up rule is employed and the experiment ends after 13 reversal points. The tactile acuity threshold is 1.6 mm is the median value of the grating widths of the last 10 reversal points. The 2-down and 1-up staircase rule converges to a 71% correct threshold25.
Figure 1: Schematic diagram illustrating the setup components for measurement of vibrotactile thresholds (A) and a typical test trial (B). Please click here to view a larger version of this figure.
Figure 2: Physical characteristics of the vibration stimulus. (A) schematic presentation of the stimulus waveform. (B) amplitudes for vibration stimuli (the x-axis has a logarithmic scale). The black circle designates the starting amplitude strength for vibration testing at 125 Hz. Please click here to view a larger version of this figure.
Figure 3: Relationship of the driving input voltage to output displacement of the piezoelectric actuator at 10 Hz and 125Hz sinusoidal input waveform, (A and B), measured by the SGS integrated sensor. (C) the relationship between the 125 Hz driving input voltage and measured output displacements for the set of vibration intensity levels used in testing. (D) Nearly identical input voltage to output displacement at the lowest intensity level, 20 nm. Please click here to view a larger version of this figure.
Figure 4: A typical performance of a subject is shown on the two-interval forced-choice vibration detection test. (A) the responses of the subject (correct, open circles; incorrect, black circles) are plotted against the trial series number. (B) a diagram depicting the adaptive method that searches for the threshold by changing both step sizes and direction (i.e., increasing and decreasing level) across a set of trials. Please click here to view a larger version of this figure.
Figure 5: Typical performance of a subject on the two-alternative forced-choice algorithm for the grating orientation task using the tactile acuity cube. The responses of the subject (correct, open circles; incorrect, black circles) are plotted against the trial number. Please click here to view a larger version of this figure.
Stimulus levels and the corresponding output voltages (V) for stimulus amplitudes | |||||
(1 V = 10 µm displacement amplitude) | |||||
1 | 0.00179 | 13 | 0.02839 | 25 | 0.45 |
2 | 0.00226 | 14 | 0.03574 | 26 | 0.56652 |
3 | 0.00284 | 15 | 0.045 | 27 | 0.7132 |
4 | 0.00357 | 16 | 0.05665 | 28 | 0.89787 |
5 | 0.0045 | 17 | 0.07132 | 29 | 1.13035 |
6 | 0.00567 | 18 | 0.08979 | 30 | 1.42302 |
7 | 0.00713 | 19 | 0.11303 | 31 | 1.79148 |
8 | 0.00898 | 20 | 0.1423 | 32 | 2.25534 |
9 | 0.0113 | 21 | 0.17915 | 33 | 2.83931 |
10 | 0.01423 | 22 | 0.22553 | 34 | 3.57448 |
11 | 0.01791 | 23 | 0.28393 | 35 | 4.5 |
12 | 0.02255 | 24 | 0.35745 |
Table 1: Stimulus levels and the corresponding output voltages (V) for stimulus amplitudes.
The techniques used to evaluate VDT vary as to device specifications, hardware, and testing protocols. The International Organization for Standardization specifies the methods and procedures to analyze and interpret vibrotactile thresholds including recommendations for the various components of a vibrometer (ISO 13091-1 and 226,27) The described testing system abides to the relevant ISO recommendations for testing frequency range (4-125 Hz), method (variant of up-down staircase and forced choice), probe size (smooth edged flat and circular), and testing at room temperature. The setup can be equipped with a sensor determining the probe's position and displacement and optionally can be equipped with a firm surround isolating the vibration stimulus.
The benefits of the described device lie in the following characteristics: availability and easy assembly of the components; adjustable frequency range 1-500 Hz; wide amplitude range (5 nm-90 µm displacement); multiple probe sizes; programmability of the vibration stimulus size and duration; and versatility to psychophysical testing protocol with various adaptive method. Although portable systems are available28, the setup is transportable and has been used in various settings such as schools, research institutes, and hospitals. Research centers, however, should have a well-equipped workshop to fabricate additional tools and devices needed for the experimental setup. There are commercially assembled and ready for use systems e.g. case IV or Medoc, but these systems are configured to deliver only a limited set of stimuli. In contrast, our system can be used to implement a very wide range of stimulation protocols that are only limited by the specifications of the piezo element used. Due to the brass bar size, it is not feasible to test other areas than the fingers and toes. Our device does not use a surround to limit the stimulation to the tested skin area and it is likely that a wider area larger than the actual skin to probe contact is stimulated. Lastly, the procedure requires attention from the subject for a longer period of time, 15 min on average. Usually the first 3 reversals have a larger change in intensity levels (4 levels) and once completed the threshold most likely lies within this range of 4 intensity levels. With progress of the test, fine determination of the threshold is made. There are other stopping criterion which could be implemented and integrated in the adaptive procedure described elsewhere29.
Several device parameters that can affect vibration detection threshold are adjustable in the setup17. These include device parameters such as the contact area of the probe to skin, use of surround encircling the probe to limit the area of stimulation, the selected vibration frequency, wrist posture, and psychophysical testing algorithm. Psychophysical testing algorithms incorporating the method of limits and the staircase method have been used to determine the VDT and there is no consensus favoring a particular method. Vibrotactile thresholds do vary depending on the psychophysical testing method used20. Estimates of VDT for 125 Hz are reliably obtained in the range of 50 to 600 nm using the outlined adaptive method in the psychophysical protocol and are in accord with VDT from other studies5,23. Furthermore; scripts for different adaptive methods can be easily developed and integrated to execute the psychophysical test for vibration detection threshold25. The adaptive procedure we described for the determination of the vibration threshold assures that for each stimulus strength level the subject's correct responses are over 76% correct in order to move down the stimulus strength staircase22. Since there are only 2 possible responses, a series of lucky guesses could erroneously alter threshold measurements, especially at low stimulus levels. For this reason we added a modification which is a series of trials at each vibration level in order to minimize such errors. Subjects who show no consistency during trial series – deviations of more than 4 levels between reversal points – are typically excluded from the study. The skin to probe contact and how the probe behaves while stimulating the skin are very important to any psychophysical experiment on touch sensation30,31. The piezo actuator is equipped with a compact strain gauge sensor (SGS) has a relatively high bandwidth (up to 3 KHz) and a very good resolution with good repeatability (0.1% of nominal displacement). Therefore, the piezoelectric device has high reliability characteristics especially when it comes to fine indentation even at high static loads. The motion is straight because the piezo (PICMA stack piezo linear actuator) we use is embedded in a guide and this ensures no lateral motion. In addition, the servo-controller can automatically compensate for varying loads or forces.
The assessment of tactile acuity we described here relies on manual delivery of tactile stimuli. The test requires careful application of the stimulus to produce perpendicular deformation of the skin and no shearing distortions that might provide cues to the subject. We choose a slightly different procedure for the determination of tactile spatial acuity and vibrotactile threshold. We did not elect to use a larger step size change initially using the TAC because it is equipped with few levels for grating width (6 levels) and these are not of constant or fixed size but vary from larger to smaller step size change. The change in size of the grating widths between the first 3 levels is 1.5 mm, between the 3rd and 4th level 1.25 mm, and 4th to 6th 0.5 mm. The subject's performance in the grating orientation task is affected by the depth of the indentation caused by the tactile stimulus, the force applied, and finger the size of the finger32,33. There are other alternatives to the tactile acuity cube: JVP domes, and the two-point discrimination task. The JVP domes are other alternatives to the tactile cube. The advantages are that the JVP domes have 8 grating widths ranging from 0.35 mm to 3.0 mm. JVP domes can be utilized to assess tactile spatial acuity of the tongue and lips8, whereas 2 point discrimination task does not rely on identifying grating orientation and is not a valid measure for tactile spatial acuity34. Recently, difficulties associated with manual testing have been improved through the introduction of an automated tactile acuity system that applies a forced-choice paradigm to determine tactile threshold for grating orientation35.
The methods for obtaining the vibrotactile and tactile acuity thresholds we described have been employed to screen individuals for touch-related traits3,7,9. In a study conducted by our group, we showed that touch traits are heritable, and that certain genetic lesions causing hearing impairment also influenced touch sensitivity7. Moreover, touch sensitivity can be assessed under different experimental condition such as water-induced fingertip wrinkling9. It will be important to make quantitative measurements of VDT and tactile acuity in patients with possible function altering mutations in genes involved in regulating mechanoreceptor sensitivity3,7. Indeed there has been a recent surge in the identification of genes directly involved in this process like Piezo2 and STOML31,13,14,36,37, and rapid progress in this area is sure to identify new genes that regulate touch. The influence of genetic variants in such "touch" genes will be ideally tested in genotyped patients with quantitative psychophysical methods like those described here.
The authors have nothing to disclose.
The research and development leading to the above described vibration test was funded by the European Research Council (ERC, ADG 294678) and the German Research Council (grant, SFB665). Thanks go to the Max-Delbrück-Center technical and support staff, Mr. R. Fischer, B. Neumann, and Mr. M. Pflaume, who provided invaluable assistance with the project.
Piezo actuator | Physik Instrument, Germany | P-602.1L | The linear piezoelectric actuator, with integrated position sensor and motion amplifier, contains a piezoceramic material that elongates and contracts when voltage is applied. The piezoelectric actuator travels up to100 µm. The actuator is equipped with a flexure guide that ensures straight motion without tilting or lateral offset. The displacement is linear and calibration is done and checked by the manufacturer. It is recommended that on-axis movement of the probe be checked under the microscope. According to the manufacturer, the stimulus amplitude dampens by less than 20% at oscillating frequencies of 1000 Hz. This can be checked by using a force or displacement measuring device (e.g. force transducer from Kleindiek). | The linear piezoelectric actuator, with integrated position sensor and motion amplifier, contains a piezoceramic material that elongates and contracts when voltage is applied. The piezoelectric actuator travels up to100 µm. The actuator is equipped with a flexure guide that ensures straight motion without tilting or lateral offset. The displacement is linear and calibration is done and checked by the manufacturer. It is recommended that on-axis movement of the probe be checked under the microscope. According to the manufacturer, the stimulus amplitude dampens by less than 20% at oscillating frequencies of 1000 Hz. This can be checked by using a force or displacement measuring device (e.g. force transducer from Kleindiek). |
Piezo Amplifier / Servo Controller | Physik Instrument, Germany | E-665 | E-665 amplifier/controller drives and controls the displacement of a low-voltage piezoelectric actuator in a system with sensor position feedback (SGS sensors). The servo-controller provides the option for closed loop operation. When applying sinusoidal and oscillating stimuli the amplitude signal deviates from the set amplitude starting from 500 Hz and reaches a maximum decrease of 20% at 1000Hz. | E-665 amplifier/controller drives and controls the displacement of a low-voltage piezoelectric actuator in a system with sensor position feedback (SGS sensors). The servo-controller provides the option for closed loop operation. When applying sinusoidal and oscillating stimuli the amplitude signal deviates from the set amplitude starting from 500 Hz and reaches a maximum decrease of 20% at 1000Hz. |
LabChart Software | ADInstruments, USA | LabChart 7, MLU60/8 | Can create, store and run macro of the psychophysical testing algorithm. | Can create, store and run macro of the psychophysical testing algorithm. |
PowerLab | ADInstruments, USA | PowerLab 4/35 PL3504 | Data Acquisition Hardware. Used with LabChart software. | Data Acquisition Hardware. Used with LabChart software. |
Brass bar | Custom-made | Bar made of pure brass, weighs 15.5Kg. When the peizoelectric actuator is mounted on the brass bar it should exert a force of 30 g weight on skin surface. | ||
Monitor | Custom-made | To mark the 1st and the 2nd interval. The monitor indicates to the subject the time intervals during which the stimulus may be presented. | ||
Response box | Custom-made | The subject indicates the interval at which stimulus occurred. | ||
Board | Custom-made | Upper surface should be smooth (Plastic), lower surface made of foam to prevent stray vibration ot be transmitted to the stimulating pobe. | ||
Probe | Custom-made | A flat circular probe with smoothed edges (thermoplastic material) attached to a screw head. The screw should be of appropriate size to be tightened directly to the moving part of piezoelectric actuator. Size of the probe can be according to preference; in our case, diameter 8.21mm and surface area 52.9mm2. | ||
Labchart Script | Can be sent on request. See supplementary code file. | |||
Tactile Acuity Cube | MedCore | The cube is comprised of 6 sides each containing a grating (bar and groove) whose widths are 0.75, 1.25, 1.75, 3.0, 4.5, and 6.0 mm. |