Physical determinants of the shape of the psychophysical curve relating tactile roughness to raised-dot spacing: implications for neuronal coding of roughness

Cognitive map formation in the blind is enhanced by three-dimensional tactile information

For blind individuals, tactile maps are useful tools to form cognitive maps through touch. However, they still experience challenges in cognitive map formation and independent navigation. Three-dimensional (3D) tactile information is thus increasingly being considered to convey enriched spatial information, but it remains unclear if it can facilitate cognitive map formation compared to traditional two-dimensional (2D) tactile information. Consequently, the present study investigated the impact of the type of sensory input (tactile 2D vs. tactile 3D vs. a visual control condition) on cognitive map formation. To do so, early blind (EB, n = 13), late blind (LB, n = 12), and sighted control (SC, n = 14) participants were tasked to learn the layouts of mazes produced with different sensory information (tactile 2D vs. tactile 3D vs. visual control) and to infer routes from memory. Results show that EB manifested stronger cognitive map formation with 3D mazes, LB performed equally well with 2D and 3D tactile mazes, and SC manifested equivalent cognitive map formation with visual and 3D tactile mazes but were negatively impacted by 2D tactile mazes. 3D tactile maps therefore have the potential to improve spatial learning for EB and newly blind individuals through a reduction of cognitive overload. Installation of 3D tactile maps in public spaces should be considered to promote universal accessibility and reduce blind individuals' wayfinding deficits related to the inaccessibility of spatial information through non-visual means.

The neural basis of tactile texture perception

Running our fingers across a textured surface gives rise to two types of skin deformations, each transduced by different tactile nerve fibers. Coarse features produce large-scale skin deformations whose spatial configuration is reflected in the spatial pattern of activation of some tactile fibers. Scanning a finely textured surface elicits vibrations in the skin, which in turn evoked temporally patterned responses in other fibers. These two neural codes-spatial and temporal-drive a spectrum of neural response properties in somatosensory cortex: At one extreme, neurons are sensitive to spatial patterns and encode coarse features; at the other extreme, neurons are sensitive to vibrations and encode fine features. While the texture responses of nerve fibers are dependent on scanning speed, those of cortical neurons are less so, giving rise to a speed invariant texture percept. Neurons in high-level somatosensory cortices combine information about texture with information about task variables.

Multiple Spatial Spectral Components of Static Skin Deformation for Predicting Macroscopic Roughness Perception

A previous study suggested a relationship between the spatial spectrum of finger pad skin deformation and perception of macroscopic roughness features. This study tested a new hypothesis that macroscopic roughness perception is the result of a weighted linear combination of multiple spatial spectral components of skin deformation. Experiments were conducted by capturing close-up images of finger pad deformation while the pads were pushed onto specimens with macroscopic features. Additionally, the roughness perceptions of these specimens were collected using a magnitude estimation method. The combination of spectral components predicted the roughness perception more accurately than any single spectral component. This suggests that roughness perception is mediated by multiple Gabor filter-like neural systems with different spatial periods, such as visual perception.

Global surface features contribute to human haptic roughness estimations

Previous studies have paid special attention to the relationship between local features (e.g., raised dots) and human roughness perception. However, the relationship between global features (e.g., curved surface) and haptic roughness perception is still unclear. In the present study, a series of roughness estimation experiments was performed to investigate how global features affect human roughness perception. In each experiment, participants were asked to estimate the roughness of a series of haptic stimuli that combined local features (raised dots) and global features (sinusoidal-like curves). Experiments were designed to reveal whether global features changed their haptic roughness estimation. Furthermore, the present study tested whether the exploration method (direct, indirect, and static) changed haptic roughness estimations and examined the contribution of global features to roughness estimations. The results showed that sinusoidal-like curved surfaces with small periods were perceived to be rougher than those with large periods, while the direction of finger movement and indirect exploration did not change this phenomenon. Furthermore, the influence of global features on roughness was modulated by local features, regardless of whether raised-dot surfaces or smooth surfaces were used. Taken together, these findings suggested that an object’s global features contribute to haptic roughness perceptions, while local features change the weight of the contribution that global features make to haptic roughness perceptions.

Effects of receptor density on the tactile perception of roughness: implications for neural mechanisms of texture perception

Aim of the study: The purpose of this study was to investigate the effects of receptor density in the glabrous skin of the hand on the perception of the roughness of a textured surface.Materials and methods: This was done by having observers make magnitude estimates of the perceived roughness of raised-dot surfaces at the fingertip, with its high receptor density, and the thenar eminence, with its much lower receptor density.Results: Judgments of perceived roughness averaged over the inter-dot spacings (0.8-5.9 mm) employed in the study did not differ significantly between the two sites, which suggested that roughness perception is not exclusively dependent upon a neural code involving variation in the activity levels of the nerve fibers of spatially distributed receptors, as is the case in spatial discrimination tasks such as spatial-gap detection, grove-orientation discrimination and letter recognition. This hypothesis was further supported by the finding that the elimination of temporal cues by preventing movement of the skin over the raised-dot surface drastically impaired judgments of perceived roughness at the thenar but had little effect on judgments of perceived roughness at the fingertip.Conclusion: These findings suggested that the neural code for perceived roughness at the fingertip is mediated primarily by spatial variation in the activity levels of spatially distributed receptors whereas the neural code for perceived roughness at the thenar is mediated primarily by temporal variation in the activity levels of individual receptors.

Relationship Between Spatial Variations in Static Skin Deformation and Perceived Roughness of Macroscopic Surfaces

The perceived roughness of macroscopic surface features is related to finger skin deformation. For example, skin indentation into surface grooves and the spatial variation in slowly adapting type 1 (SAI) mechanoreceptor activities are linked to the magnitude of subjective roughness. However, the perception of macroscopic roughness has not been directly linked to the spatial variations in skin deformation; therefore, this article investigated the relationship between the subjective roughness magnitude and the spatial spectrum of skin deformation through contact with roughened macroscopic surfaces. Experiments were performed to measure deformation of a finger pad when it statically touched different shaped grating scales. Then, the spatial spectra of skin deformation were computed by applying the Gabor filter with varying spatial selectivity. Some spectral components, particularly those for the spatial period ranging from 2.45-4.00 mm, exhibited a good correlation with the perceived roughness magnitude. When the optimal spatial periods were determined for individual subjects, the spectral components and perceived roughness exhibited a strong linear relationship with an average correlation coefficient of 0.94. Therefore, the amount of spatial variation in finger skin deformation can be linked to the subjective roughness intensities of macroscopic surface features.

The neural code for tactile roughness in the somatosensory nerves

Roughness is the most salient perceptual dimension of surface texture but has no well-defined physical basis. We seek to determine the neural determinants of tactile roughness in the somatosensory nerves. Specifically, we record the patterns of activation evoked in tactile nerve fibers of anesthetized Rhesus macaques to a large and diverse set of natural textures and assess what aspect of these patterns of activation can account for psychophysical judgments of roughness, obtained from human observers. We show that perceived roughness is determined by the variation in the population response, weighted by fiber type. That is, a surface will feel rough to the extent that the activity varies across nerve fibers and varies across time within nerve fibers. We show that this variation-based neural code can account not only for magnitude estimates of roughness but also for roughness discrimination performance.NEW & NOTEWORTHY Our sense of touch endows us with an exquisite sensitivity to the microstructure of surfaces, the most salient aspect of which is roughness. We analyze the responses evoked in tactile fibers of monkeys by natural textures and compare them to judgments of roughness obtained for the same textures from human observers. We then describe how texture signals from three populations of nerve fibers are integrated to culminate in a percept of roughness.

A Variation Code Accounts for the Perceived Roughness of Coarsely Textured Surfaces

For decades, the dominant theory of roughness coding in the somatosensory nerves posited that perceived roughness was determined by the spatial pattern of activation in one population of tactile nerve fibers, namely slowly adapting type 1 (SA1) afferents. Indeed, the perceived roughness of coarsely textured surfaces tracks the spatial variation in SA1 responses – the degree to which response strength varies across SA1 afferents. However, in a later study, the roughness of a different set of dot patterns was found to be a monotonic function of dot spacing, a result interpreted as evidence that roughness was determined by the strength of SA1 responses – the population firing rate – rather than their spatial layout. Then again, the spatial variation hypothesis was not tested directly as afferent responses to the conflicting patterns were not measured. To fill this gap, we simulated afferent responses to the dot patterns used in these roughness coding experiments using a model of skin mechanics. We then implemented the spatial variation and firing rate models of roughness based on these simulated responses to generate predictions of perceived roughness. We found that the spatial variation model accounts for perceived roughness under all tested conditions whereas the firing rate model does not.

Tactile texture signals in primate primary somatosensory cortex and their relation to subjective roughness intensity

This study investigated the hypothesis that a simple intensive code, based on mean firing rate, could explain the cortical representation of subjective roughness intensity and its invariance with scanning speed. We examined the sensitivity of neurons in the cutaneous, finger representation of primary somatosensory cortex (S1) to a wide range of textures [1 mm high, raised-dot surfaces; spatial periods (SPs), 1.5-8.5 mm], scanned under the digit tips at different speeds (40-115 mm/s). Since subjective roughness estimates show a monotonic increase over this range and are independent of speed, we predicted that the mean firing rate of a subgroup of S1 neurons would share these properties. Single-unit recordings were made in four alert macaques (areas 3b, 1 and 2). Cells whose discharge rate showed a monotonic increase with SP, independent of speed, were particularly concentrated in area 3b. Area 2 was characterized by a high proportion of cells sensitive to speed, with or without texture sensitivity. Area 1 had intermediate properties. We suggest that area 3b and most likely area 1 play a key role in signaling roughness intensity, and that a mean rate code, signaled by both slowly and rapidly adapting neurons, is present at the level of area 3b. Finally, the substantial proportion of neurons that showed a monotonic change in discharge limited to a small range of SPs (often independent of response saturation) could play a role in discriminating smaller changes in SP.

The tactile speed aftereffect depends on the speed of adapting motion across the skin rather than other spatiotemporal features

After prolonged exposure to a surface moving across the skin, this felt movement appears slower, a phenomenon known as the tactile speed aftereffect (tSAE). We asked which feature of the adapting motion drives the tSAE: speed, the spacing between texture elements, or the frequency with which they cross the skin. After adapting to a ridged moving surface with one hand, participants compared the speed of test stimuli on adapted and unadapted hands. We used surfaces with different spatial periods (SPs; 3, 6, 12 mm) that produced adapting motion with different combinations of adapting speed (20, 40, 80 mm/s) and temporal frequency (TF; 3.4, 6.7, 13.4 ridges/s). The primary determinant of tSAE magnitude was speed of the adapting motion, not SP or TF. This suggests that adaptation occurs centrally, after speed has been computed from SP and TF, and/or that it reflects a speed cue independent of those features in the first place (e.g., indentation force). In a second experiment, we investigated the properties of the neural code for speed. Speed tuning predicts that adaptation should be greatest for speeds at or near the adapting speed. However, the tSAE was always stronger when the adapting stimulus was faster (242 mm/s) than the test (30-143 mm/s) compared with when the adapting and test speeds were matched. These results give no indication of speed tuning and instead suggest that adaptation occurs at a level where an intensive code dominates. In an intensive code, the faster the stimulus, the more the neurons fire.

The Effect of Task Instruction on Haptic Texture Processing: The Neural Underpinning of Roughness and Spatial Density Perception

Perceived roughness is associated with a variety of physical factors and multiple peripheral afferent types. The current study investigated whether this complexity of the mapping between physical and perceptual space is reflected at the cortical level. In an integrative psychophysical and imaging approach, we used dot pattern stimuli for which previous studies reported a simple linear relationship of interdot spacing and perceived spatial density and a more complex function of perceived roughness. Thus, by using both a roughness and a spatial estimation task, the physical and perceived stimulus characteristics could be dissociated, with the spatial density task controlling for the processing of low-level sensory aspects. Multivoxel pattern analysis was used to investigate which brain regions hold information indicative of the level of the perceived texture characteristics. While information about differences in perceived roughness was primarily available in higher-order cortices, that is, the operculo-insular cortex and a ventral visual cortex region, information about perceived spatial density could already be derived from early somatosensory and visual regions. This result indicates that cortical processing reflects the different complexities of the evaluated haptic texture dimensions. Furthermore, this study is to our knowledge the first to show a contribution of the visual cortex to tactile roughness perception.

Natural scenes in tactile texture

Sensory systems are designed to extract behaviorally relevant information from the environment. In seeking to understand a sensory system, it is important to understand the environment within which it operates. In the present study, we seek to characterize the natural scenes of tactile texture perception. During tactile exploration complex high-frequency vibrations are elicited in the fingertip skin, and these vibrations are thought to carry information about the surface texture of manipulated objects. How these texture-elicited vibrations depend on surface microgeometry and on the biomechanical properties of the fingertip skin itself remains to be elucidated. Here we record skin vibrations, using a laser-Doppler vibrometer, as various textured surfaces are scanned across the finger. We find that the frequency composition of elicited vibrations is texture specific and highly repeatable. In fact, textures can be classified with high accuracy on the basis of the vibrations they elicit in the skin. As might be expected, some aspects of surface microgeometry are directly reflected in the skin vibrations. However, texture vibrations are also determined in part by fingerprint geometry. This mechanism enhances textural features that are too small to be resolved spatially, given the limited spatial resolution of the neural signal. We conclude that it is impossible to understand the neural basis of texture perception without first characterizing the skin vibrations that drive neural responses, given the complex dependence of skin vibrations on both surface microgeometry and fingertip biomechanics.

Neuronal correlates of tactile speed in primary somatosensory cortex

Moving stimuli activate all of the mechanoreceptive afferents involved in discriminative touch, but their signals covary with several parameters, including texture. Despite this, the brain extracts precise information about tactile speed, and humans can scale the tangential speed of moving surfaces as long as they have some surface texture. Speed estimates, however, vary with texture: lower estimates for rougher surfaces (increased spatial period, SP). We hypothesized that the discharge of cortical neurons playing a role in scaling tactile speed should covary with speed and SP in the same manner. Single-cell recordings (n = 119) were made in the hand region of primary somatosensory cortex (S1) of awake monkeys while raised-dot surfaces (longitudinal SPs, 2-8 mm; periodic or nonperiodic) were displaced under their fingertips at speeds of 40-105 mm/s. Speed sensitivity was widely distributed (area 3b, 13/25; area 1, 32/51; area 2, 31/43) and almost invariably combined with texture sensitivity (82% of cells). A subset of cells (27/64 fully tested speed-sensitive cells) showed a graded increase in discharge with increasing speed for testing with both sets of surfaces (periodic, nonperiodic), consistent with a role in tactile speed scaling. These cells were almost entirely confined to caudal S1 (areas 1 and 2). None of the speed-sensitive cells, however, showed a pattern of decreased discharge with increased SP, as found for subjective speed estimates in humans. Thus further processing of tactile motion signals, presumably in higher-order areas, is required to explain human tactile speed scaling.