The human brain processes ongoing sensory information by comparing incoming sensory information to previous stimulation, which enables humans to detect changes in their environment (Laufer et al., 2008). As humans, we explore our haptic environment through active touch and the glabrous skin on our hands and digits (Gibson, 1962; Wagner and Gibson, 2016). Active touch is integral in identifying and evaluating objects and surfaces by optimising contact pressure, speed, and velocity (Lederman and Klatzky, 2009). The present study aimed to investigate how the brain encodes continuous tactile information associated with changes in touch experience by assessing neural oscillations related to perceptual texture change.

Processing of stimulus change in the brain has previously been investigated using electroencephalography (EEG) and oddball tasks (Naatanen et al., 1978), wherein participants are exposed to one repetitive stimulus and then presented with a novel oddball stimulus. This type of paradigm results in an event-related potential (ERP) response called mismatch negativity (MMN); a correlate of auditory change perception that peaks around 100–300 ms following novel stimuli (Näätänen et al., 2005, 2007). Though primarily studied in the auditory domain, MMN effects have been reported in other sensory modalities such as vision (Stefanics et al., 2014) and somatosensation (sMMN; Restuccia et al., 2007; Butler et al., 2011, 2012; Hu et al., 2013; Chen et al., 2014). Oddball paradigms present stimuli for fixed intervals with a period of rest prior to stimulation, thus revealing information about the brain processing of individual events and features. A drawback of this approach is that this type of stimulation is not representative of real-world experiences, where ongoing sensory processing occurs, and change detection is an additive experience rather than an isolated feature.

In contrast to ERP analysis, investigating neural oscillations using time-frequency analysis provides a more accurate representation of ongoing sensory processing in the brain, by offering insights into neuronal synchrony and ongoing inter-neuronal communication (Mathalon and Sohal, 2015; Morales and Bowers, 2022). This approach enables the investigation of the summation of neural oscillations and ongoing changes in the brain in response to external stimuli (Cohen, 2014). Alpha-band and beta-band oscillations are attenuated during tactile stimulation or voluntary movement (Chatrian et al., 1958; Pfurtscheller, 1981; Salmelin and Hari, 1994) over the primary somatosensory and motor cortices, respectively (Brovelli et al., 2004). Attenuation of alpha- and beta-band oscillations are thought to reflect increased cortical activation, whereas the presence of synchronous oscillations is indicative of cortical areas at rest (Pfurtscheller, 1992, 1999).

Texture processing in the brain manifests bilaterally as attenuation of alpha-and beta-band rhythms relative to a rest period (Genna et al., 2018; Henderson et al., 2022). Whilst active exploration of texture has been investigated with EEG (Henderson et al., 2022), the neural mechanisms that underpin the processing of texture change during naturalistic explorations are relatively unknown. Processing of rough and smooth textures demonstrate altered cortical responses, with smooth textures typically eliciting increased brain activation relative to rougher surfaces during passive tactile stimulation (Moungou et al., 2016; Genna et al., 2018). Comparatively, our recent study investigating the neural correlates of texture perception with active touch found increased attenuation of beta-band oscillations for smooth textures and increased attenuation of alpha-band oscillations for rough textures (Henderson et al., 2022). Taken together, the literature indicates that cortical changes which occur during texture processing are modulated by textural properties, such as surface roughness. Therefore, change detection in the brain during ongoing processing is also likely to demonstrate altered neural oscillatory response as an indication of encoding textural change.

Change detection in the brain has been previously investigated using time-frequency analysis of MMN ERPs, which revealed increased theta-band power across auditory (Fuentemilla et al., 2008; Hsiao et al., 2009; Ko et al., 2012), visual (Stothart and Kazanina, 2013; Liang et al., 2017), and somatosensory modalities during unattended pressure stimulation to the finger pad (Zhang et al., 2019). Theta-band oscillations have also been linked with top-down memory processes, which led to increased power in frontal midline regions (Klimesch, 1999; Klimesch et al., 2008). Further, ERPs in the time-domain are suggested to manifest as theta-band changes in the time-frequency domain (Bernat et al., 2007; Harper et al., 2014). Therefore, theta-band power may reflect event-related changes in neural processing of texture change. However, it should be noted that findings from oddball tasks are not representative of complex environmental change, and therefore oscillatory brain activity may differ under active ongoing exploration.

The present study utilised touch sensor technology to investigate cortical oscillatory changes in alpha-, beta- and theta-bands during active touch of two adjacent textures. A touch sensor was used to quantify touch behaviour in real-time and compute time markers from when the index finger crossed the point of texture transition. Time markers were subsequently integrated with EEG data to consider event-related changes in oscillatory activity. We hypothesised that brain oscillations would show differences related to texture change; specifically, there will be reductions in alpha- and beta-band power over sensorimotor areas related to perception of texture change. Further, we hypothesise that theta-band power would increase as an oscillatory reflection of perceptual change mechanisms in frontal regions. A secondary analysis compared texture change from rough-to-smooth, relative to smooth-to-rough transitions. We hypothesised beta-band power would decrease over sensorimotor areas when transitioning from rough to smooth, whereas alpha-band power would decrease when transitioning from smooth to rough, demonstrating texture differences in frequency bands in line with our previous research (Henderson et al., 2022).

The present study aimed to establish how the brain encodes texture change by assessing oscillatory differences in alpha-, beta- and theta-bands as the finger transitions from one texture to another during active touch exploration. Texture transition produced significant differences in alpha and beta frequency bands, whereas no statistically significant effects were observed in theta-band. In line with our hypothesis, alpha-band power over sensorimotor cortical regions decreased during texture transition when compared to pre-transition texture processing. Further, beta-band power decreased during texture transition when transitioning from rough-to-smooth textures over central regions, relative to smooth-to-rough trials, supporting previous findings from our lab (Henderson et al., 2022). Results indicate that perceptual texture change is observable in oscillatory brain activation patterns recorded by EEG under a naturalistic paradigm.

Bilateral decreases in alpha-band power were observed for texture transition relative to pre-transition texture processing over sensorimotor regions. Alpha-band power is associated with bilateral activation following tactile stimulation (Tamè et al., 2016; Genna et al., 2018), wherein a decrease in power, or attenuation of oscillations, signifies an increase in cortical processing (Cheyne et al., 2003; Gaetz and Cheyne, 2006). The present results indicate that it is likely that alpha-band oscillations also reflect perceptual texture change mechanisms in the brain, which manifests as increased cortical activation (decrease in power) to encode novel or changing tactile input from mechanoreceptors. Previous research with active touch has suggested that alpha-band power is also associated with roughness, wherein rougher textures increase cortical activation (Henderson et al., 2022), whilst passive touch research reports increased cortical activation for smoother textures (Ballesteros et al., 2009; Moungou et al., 2016; Genna et al., 2018). In contrast to our hypothesis, the present study demonstrated no significant effect in alpha-band when comparing rough to smooth transition and vice-versa. This suggests that the effects found in alpha-band for this study are representative of change detection, rather than roughness modulation. Therefore, it is possible that the role of alpha-and oscillations during change detection differs from that during roughness encoding in tactile stimulation, though further research is necessary to confirm this hypothesis.

Beta-band power was decreased over central sensorimotor regions when comparing transition from rough to smooth with the opposite condition. This supports our hypothesis that smooth textures increase cortical activation in the beta-band and accords with previous studies using passive (Moungou et al., 2016; Genna et al., 2018) as well as active touch paradigms (Henderson et al., 2022). The duplex theory states that the perception of fine textures is mediated by high-frequency vibrations from tactile elements (Katz, 1925, 1989; Hollins and Risner, 2000). Previous research suggests that beta-band activity is mediated by vibration intensity (Park et al., 2021), which contributes to observed difference in the EEG signal between rough and smooth textures (Henderson et al., 2022). Interestingly, the present study demonstrated that texture modulates beta-band power during transitioning between two textures, which, to the best of our knowledge, is the first instance demonstrating this effect.

Increased frontal theta-band oscillations have previously been shown to coincide with the enhancement of the sMMN component (Zhang et al., 2019). We hypothesised that such components may be visible in theta-band changes in time-frequency analysis. However, in this naturalistic active touch paradigm, perceptual texture change did not modulate theta-band power per se. Therefore, identification of change detection ERP components may not be detectable with the current temporal resolution (100 Hz) computation of time markers. Time-locking may need to be more precise to directly compare well known change perception ERPs, such as the sMMN, in ongoing time-frequency changes. However, this is the first known study to investigate change perception in active touch in a naturalistic paradigm, offering a more ecologically valid approach than previous oddball paradigms with stationary participants. Therefore, it is also possible that ERP correlates of change detection seen in previous studies differ, or are not present, when performing ongoing explorations of texture.

The present study used continuous tactile stimulation to investigate perceptual texture change in the way that we typically encounter it in our environment, as change occurring on top of an ongoing sensory experience (Gallace et al., 2006, 2007). Contrasting with rest improves the signal-to-noise ratio of post-stimulus activity (Toro et al., 1994). There was no rest period included in this paradigm, which incidentally reduces the signal-to-noise ratio and results in small effect sizes due to the comparison of brain oscillations with active baseline conditions. However, to understand the features of change detection, it was essential to compare texture change to ongoing tactile perception. Previous EEG research has investigated perceptual change detection using complex ongoing auditory (Boubenec et al., 2017) and visual stimuli (O’Connell et al., 2012; Kelly and O’Connell, 2013). Therefore, ongoing stimulation is appropriate for perceptual change paradigms, as they allow researchers to understand the complexities of sensory experience during more naturalistic tasks. Though future paradigms may also benefit from a period of pre-stimulus rest to improve the signal-to-noise ratio for additional analyses in tandem with an active baseline approach.

There are limitations to the present design. As a lab-based study, and due to the trial requirements for time-frequency EEG analysis (Cohen, 2016), participants were exposed to the two textured stimuli repeatedly over the course of the experiment. Therefore, it is likely that changes in texture were predictable, which may have developed expectations and diminished stimulus novelty. Further, repeated stimulation may have led to sensory desensitisation (Klingner et al., 2011; Graczyk et al., 2018) and reduced task engagement (Lelis-Torres et al., 2017). However, repeated trials are necessitated by the time-frequency method (Cohen, 2016). Future research could consider the use of one repetitive stimulus embedded within a more complex paradigm with various novel stimuli and various gradations of intensity to manipulate the degree to which the textural change was detectable whilst maintaining novelty, engagement and reducing desensitisation.

In conclusion, the present study demonstrates that alpha-band activity is related to perceptual texture change during continuous texture exploration, whilst beta-band may be linked to the processing of vibrotactile cues. Therefore, this study hypotheses that alpha- and beta-band both play a functional role in acting as a change detection mechanism as well as processing surface properties of the texture, respectively (Henderson et al., 2022). However, the neural underpinnings of texture processing require further elucidation. Future research should consider using active exploration of one repetitive stimulus alongside various novel stimuli, which will enable researchers to maintain novelty and test the hypothesis that alpha-band activity encodes perceptual texture change. To our knowledge, the results demonstrate for the first time that the encoding of textural change detection can be measured in the brain during ongoing active exploration of surfaces.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found at: https://osf.io/s8xt6/?view_only=3c85a99cb2f948caa3439f2a7823b281.

The studies involving human participants were reviewed and approved by the study was approved by the Research Ethics Committee of the University of Liverpool. The patients/participants provided their written informed consent to participate in this study.

JH: conceptualization, methodology, software, formal analysis, investigation, data curation, writing–original draft, writing–review and editing, visualisation, and project administration. TM, AN-F, and DH: investigation and writing–review and editing. AH: methodology, software, and writing–review and editing. TG: conceptualization, funding acquisition, and supervision. AM and AS: conceptualization, supervision, and writing–review and editing. NF: conceptualization, methodology, software, formal analysis, writing–review and editing, supervision, and funding acquisition. All authors contributed to the article and approved the submitted version.

This work was supported by the EPSRC [Grant number: EP/S513830/1], and Unilever.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.