HHF Part4 - Tactile Sensing

Part 4 Tactile Sensing

Within the domain of tactile sensing, it is important to distinguish between tasks that involve stimulating the stationary hand with a stationary object (“passive static”) and tasks that involve an external agent moving an object across the stationary hand (“passive movement”).

요약하기보다는, 실험결과들 중 내 눈길을 끈 것들 or take-away 정리. 

Pressure

- Although Weinstein's monofilaments were calibrated in terms of force, Levin, Pearsall, and Ruderman (1978) have argued that stress (force per unit area) is the more appropriate variable to use when measuring pressure sensitivity.

- Sensitivity of the human fingerpad to tangential force is lower than that to normal force, presumably due to the high impedance of the fingerpad to tangential stimulation (Biggs & Srinivasan, 2002a). As we shall see presently, tangential forces play a role in the perception of surface texture and shape.

https://www.researchgate.net/figure/The-force-accompanying-the-finger-pad-interactions-with-a-contact-surface_fig2_380929145

Temperature

- Nevertheless, both studies confirm that the hand is more sensitive to cold than to warmth.

- Almost complete temporal summation (i.e., the summation of warmth over time) occurs on the forehead over a range of temporal durations from 0.5 to 1.0s, with no further effect up to 10s (Stevens,Okulicz,&Marks,1973). We know of no scientific studies that have parametrically assessed either temporal or spatial summation effects on the hand. Although both summation effects are likely to occur, there may well be differences between the hand and other body regions.

Electrocutaneous Stimulation

- The results are shown in figure 4.2. For frequencies ranging from 60 to 1,000 pulses/s, sensitivity (measured as the inverse of the observer's absolute threshold relative to baseline response) increases with increasing unidirectional pulse width, indicating the occurrence of temporal summation effects for electrocutaneous stimulation of the hand.

- Electrocutaneous spatial summation has also been investigated. For example, greater-than-complete summation, or supersummation, occurs at low sensation levels, whereas complete spatial summation occurs at high sensation levels (Higashiyama & Tashiro, 1990,1993a).

- Weber fraction (i.e., difference threshold/intensity of the standard stimulus

Vibration

- Two assumptions are made. First, the channel with the greatest sensitivity processes threshold sensations. Second,more than one activated psychophysical channel is responsible for producing suprathreshold sensations, although it can be as few as two. The outputs of all activated channels are presumably combined to produce a unitary sensation at cortical levels higher than area SI. Table 4.1 summarizes the essential psychophysical characteristics of each channel.

- The hand is also susceptible to vibrotactile “after-effects.” For example, when the skin is stimulated by an intense vibrotactile adapting stimulus of a given frequency for 1–2 min, the threshold for a neutral test stimulus (vibration of the same frequency) is lower than when the adapting stimulus has not been presented, demonstrating what is known as a negative aftereffect. (Sometimes, however, observers report that the test stimulus feels more intense following exposure to the adapting stimulus, that is, a positive aftereffect.)

Spatial Acuity

- Figure 4.7 shows the mean two-point thresholds for different regions on the hand averaged over the left and right sides for females (panel A) and for males (panel B) (Weinstein, 1968). The most consistent difference between the palm and the five digits is that the digits resolve spatial detail more precisely than the palm does.

- Thus, the thermal sensations experienced directly at the outer fingers were also “referred” or misdirected spatially to the middle finger, the site of the thermally neutral stimulus. A simple demonstration of the illusion may be produced by warming coins on a radiator or cooling them in a refrigerator.T he illusion is depicted in figure4.9. To summarize, when there was a spatial discrepancy between tactile and thermal patterns on the skin, the more spatially acute tactile sense dominated the less spatially acute thermal sense.

Temporal Acuity

 Audition > Tactile > Vision

- With respect to temporal resolving capacity, the visual system can determine the successiveness of two pulses separated by about 25 ms, whereas the auditory system requires only 0.01 ms (Sherrick & Cholewiak, 1986). With a corresponding value of 5 ms, mechanical touch lies in an intermediate range between vision and audition, with audition being best and vision the worst. Numerousness, that is, the ability of the observer to count accurately a series of sensory events (visual, auditory, or tactile) presented within a given time period produces a similar temporal ordering of the senses (Lechelt, 1975). Such results are likely important when considering issues pertaining to the nature of multisensory integration (see Calvert, Stein, & Spence, 2004) and the design of multimodal interfaces for teleoperation and virtual environment systems (see chapter 10)

Tactile Motion

- Smooth apparent motion, or beta motion, is detected at various sites (e.g., the back) when two tactile stimuli (optimally, periodic such as 150 Hz) are delivered to two different sites on the skin. The stimulus appears to move between the two contact sites when the interval between stimulus onset falls within a certain range.

- When the adapting stimulus is replaced with a stationary stimulus, which in the unadapted state is presumed to elicit equal activity in neurons selectively sensitive to all directions of motion, the activity-weighted mean response of the adapted cortical network is displaced away from the direction of the movement of the adapting stimulus. It has been proposed that the observer interprets this as a shift in perceived movement in a direction opposite to that of the adapting stimulus. Unfortunately, researchers have not always been able to produce the tactile aftereffect reliably, as noted both by Hollins and Favorov (1994) and more recently by Lerner and Craig (2002)

Space-Time Interactions in Somatosensory Processing

- Although this perceptual phenomenon has been confirmed with respect to audition (Cohen, Hansel, & Sylvester, 1954) and vision (Cohen, Hansel, & Sylvester, 1953), it has not been demonstrated in the tactile domain (Yoblick & Salvendy, 1970). Moreover, none of the work to date has selected the hand as a site of stimulation. The occurrence of tactile tau and kappa should be explicitly evaluated with respect to the hand.

- The tau effect refers to the fact that the apparent distance between three equally spaced stimuli that are presented successively on the forearm depends upon the intervening temporal interval (Helson&King,1931). More specifically, if the time between the first and second stimuliis shorter (longer) than the time between the second and third stimuli, people judge the corresponding distance between the first and second stimuli to be shorter (longer) than the physically equal distance between the second and third stimuli. This effect has been demonstrated in all three modalities. The converse effect, known as kappa, refers to the fact that the apparent temporal interval between two successively presented stimuli depends upon their spatial separation. Specifically, the temporal interval is perceived to increase with increasing spatial extent.

- A different but striking space-time interaction known as the saltation effect involves the illusory displacement of tactile stimuli (Geldard & Sherrick, 1972)

Saltation Effect:

• Description: The saltation effect is an illusion in which a series of tactile stimuli, such as taps or vibrations, are delivered in a specific sequence across different locations on the skin. Instead of perceiving each stimulus at its precise location, the brain integrates these stimuli and creates the sensation that they are "jumping" or "leaping" smoothly from one location to another. The perception of movement or "leap" occurs even though the stimuli are physically presented at discrete points.

 

- Studies of real and apparent motion, motion aftereffects, and space-time interactions have important consequences fo runderstanding how we normally process motion, space, and form via the skin. In addition, they are relevant to the design of tactile displays for communication and teleoperator and virtual environment applications, as discussed in chapter 10.

Properties of Surfaces and Objects

- We have previously noted that when an external agent stimulates the stationary hand with a stimulus (passive touch), observers tend to focus on their own internal sensations. In contrast, when they actively explore the stimulus (active touch), their perceptual experiences tend to be directed toward the external object and its properties(Gibson,1962)

- Weight reflects the contributions of both material and geometric features since it is partly determined by object mass, which in turn is the product of object density(a material feature) and object volume(a geometric feature).

Material Properties - Surface Texture

- It is important to understand that the perception of texture is multidimensional. 

- Moreover, with two-dimensional raised-dot patterns, people can discriminate differences in spatial period (that is, the sum of one element width and one interelement width) as small as 2% with 75% accuracy whether active or passive touch is used (Lamb, 1983).

Material Properties - Surface Roughness

- ~as research indicates that the perception of roughness is similar under both conditions provided there is relative motion between the skin and the surface (Lamb, 1983; Lederman, 1981, 1983; Verrillo, Bolanowski, & McGlone, 1999). This suggests that cutaneous (as opposed to haptic, which includes both cutaneous and kinesthetic) inputs primarily determine perceived roughness.

- On the basis of these three sets of converging results, vibration (and, therefore, temporal factors) appears to play no role in tactile roughness perception.

- What are groove, ridge, and grating?

from DALLE-2

- Collectively,the three studies suggest a single-code theory of roughness perception.

What is a single-code theory?

The single-code theory of roughness perception is a concept in tactile perception that proposes a unified explanation for how the brain interprets the sensation of roughness when we touch different surfaces. According to this theory, the perception of roughness is primarily determined by a single type of neural signal, or "code," which represents the spatial characteristics of a surface that are detected by mechanoreceptors in the skin.

Key Points of the Single-Code Theory:

1. Neural Coding: The theory suggests that the roughness of a surface is encoded by the frequency and intensity of nerve impulses generated by specific mechanoreceptors in the skin, particularly those that respond to fine spatial details, such as the SA1 (slowly adapting type 1) mechanoreceptors. These receptors are sensitive to the spatial patterns and textures on a surface.
2. Spatial Summation: Roughness perception is thought to result from the brain's ability to integrate information from multiple mechanoreceptors across a contact area. The combined input provides a single, coherent signal that the brain interprets as the level of roughness.
3. Critical Role of Spatial Features: The theory emphasizes that the spatial properties of a surface—such as the size, distribution, and spacing of surface elements (like bumps or ridges)—are key factors that influence the perceived roughness. These features directly affect how the mechanoreceptors are activated, which in turn determines the roughness signal sent to the brain.
4. Subjective Roughness: The perceived roughness is not just a direct mapping of the physical properties of a surface but is influenced by how the neural code generated by the mechanoreceptors is processed by the brain. Thus, even if two surfaces have different physical textures, they might be perceived as having similar roughness if they generate similar neural codes.

- Up to this point in the discussion, we have considered single-code models of roughness perception mediated by the population of SA I units. However, Hollins and his colleagues (e.g., Hollins et al., 1998; Hollins, Bensmaia, & Washburn, 2001; Hollins & Risner, 2000) have proposed a duplex model of tactile roughness perception that differentiates between the perceptions of fine versus coarse textures.

- Bensmaïa and Hollins offered additional support for the proposal that perception and discrimination of very fine surface textures may be mediated by the PC channel and in addition that perceive d roughness varies as a function of the vibratory intensity scaled to the spectral sensitivity of the PC system.

- The results revealed that the mean normalized root mean square of the tangential force rate changed by approximately 21% and the subjective estimates of roughness by about 16%. The collective results from the two experiments suggest that the root mean square of the rate of change in tangential force may also contribute to perceived roughness magnitude.

Material Properties - Compliance

- Cutaneous information alone proved sufficient for discriminating the rubber surfaces. Presumably, because the contact force was constant, the spatial distribution of pressure within the contact region on the fingertip could be used to differentiate objects in terms of their relative compliance

Material Properties - Thermal Properties

- These studies indicate that thermal cues can be used to recognize objects by touch, but as the resting temperature of the skin is typically higher than the ambient temperature of objects encountered in the environment, the thermal conductance and heat capacity of different materials are used to assist in identifying what the object is made from. The absolute temperature is not informative because at room temperature it is usually the same for each object palpated. The coding of object temperature is therefore primarily determined by the activity of cold receptors, which signal the cooling of the skin.

Material Properties - Weight

- The cold weights felt substantially heavier than the neutral weights at all body sites. Warm weights also tended to intensify heaviness sensations, but the results were only statistically significant on the palm, forearm, and upper arm. These results may be interpreted based on an earlier proposal offered by J. C. Stevens and Green (1978), namely, that thermal intensification of perceived heaviness is due to the failure of mechanoreceptors to respond selectively to mechanical stimulation.

Material - Geometric Properties

- Overall, force has a variable effect on perceived curvature, as compared to the highly consistent effect due to changes in physical curvature. Finally, we note that curvature discrimination of strips presented statically to the hand is best when placed along as opposed

-The well-known visual distortion effect known as the horizontal-vertical illusion has been documented with passive motion of a line stimulus across the forearm (Wong, Ho, & Ho, 1974), as well as with active touch (Day & Wong, 1971; Heller & Joyner, 1993; Millar & Al-Attar, 2000):

- For example, Loomis (1981b, 1982) demonstrated strong similarity between tactile (fingertip) and visual legibility of character sets differing in size and typography (e.g., Roman letters, Braille characters, etc.)when visual acuity was matched to tactile acuity via opticalblurring(figure4.19)

- Collectively, Loomis's results further confirm that tactile pattern recognition is spatially limited only by the spatial resolution of the cutaneous system, as determined by the spacing of the peripheral SA I mechanoreceptors (see also Johnson & Phillips, 1981; Phillips, Johansson, & Johnson, 1990).

- Craig(1982b) also confirmed the importance of temporal integration(i.e.,the time course over which temporally separated patterns can be integrated), using several different procedures. Collectively, the results suggest that the cutaneous system can totally integrate the pattern over a period lasting less than 10 ms and that the accuracy of tactile letter recognition asymptotes at about 50 ms.

-However, when traced on the backward-facing palm with the elbow bent, the percept corresponds to a normally oriented 2; in contrast, when the palm is rotated away from the body with the thumb pointing upward, a reversed 2 is perceived (figure 4.21). Note that in this second example, the percept changes despite an unchanging pattern of cutaneous stimulation. Thus, the observer must be processing the cutaneous information within some allocentric (i.e., external) frame of reference. Based on a number of informal observations, Corcoran proposed that observers view these patterns as if they were drawn on a transparent body via a “disembodied eye” positioned behind and a little above the observer.

- Their experiments further revealed that with respect to three-dimensional haptic space, observers make judgments with respect to the “above”/“below” axis in allocentric terms, with respect to the “near”/“far” axis in egocentric terms (i.e., with reference to the observer), and with respect to the left/right axis in terms of a combination of allocentric and egocentric factors.

-~spatial and temporal filtering characteristics of the tactile system, the observer is also influenced in a complex manner by the direction of the pattern that is applied relative to the body and to the external environment.