Edited by: Tetsuo Kida, National Institute for Physiological Sciences, Japan
Reviewed by: Chloe Alexandre, Beth Israel Deaconess Medical Center, USA; Elia Valentini, University of Essex, UK
*Correspondence: Enrico Schulz
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution and reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Although humans are generally capable of distinguishing single events of pain or touch, recent research suggested that both modalities activate a network of similar brain regions. By contrast, less attention has been paid to which processes uniquely contribute to each modality. The present study investigated the neuronal oscillations that enable a subject to process pain and touch as well as to evaluate the intensity of both modalities by means of Electroencephalography. Nineteen healthy subjects were asked to rate the intensity of each stimulus at single trial level. By computing Linear mixed effects models (LME) encoding of both modalities was explored by relating stimulus intensities to brain responses. While the intensity of single touch trials is encoded only by theta activity, pain perception is encoded by theta, alpha and gamma activity. Beta activity in the tactile domain shows an on/off like characteristic in response to touch which was not observed in the pain domain. Our results enhance recent findings pointing to the contribution of different neuronal oscillations to the processing of nociceptive and tactile stimuli.
The somatosensory system senses environmental stimuli (e.g., mechanical, thermal, vibrational or painful stimuli) by different types of skin receptors. Modality-specific sensory information is then conveyed from the periphery to the cerebral cortex via separate fiber tracts, e.g., the lateral spinothalamic tract for the transmission of pain and temperature (Melzack and Casey,
In the past, studies that investigated the pain modality often used somatosensory stimulation only as control condition (e.g., Seminowicz et al.,
The present EEG study aimed at investigating the differences and commonalities of pain and touch processing in the human brain. Two novel aspects will be explored. First, we will show the topographical distribution of the averaged neuronal responses to laser pain stimuli and touch stimuli. Tactile processing will be explored with an elaborated pneumatic device for delivering natural stimuli that are comparable with laser pain trials in terms of latency and duration. Second, we will investigate and compare—at single trial level—how the encoding of intensities of both modalities is subserved in the human brain. Linear mixed effects models (LME) will quantify which neuronal responses vary with different intensities of pain and touch. We aim at exploring which responses commonly or uniquely encode the intensities of pain and touch.
Nineteen healthy male human subjects with a mean age of 24 years (21–31 years) participated in the experiment, which formed the control condition of a study on the role of dopamine in pain processing (Tiemann et al.,
In two consecutive counterbalanced conditions, 75 painful cutaneous laser stimuli and 75 tactile stimuli of matched intensities were delivered to the dorsum of the right hand. The laser device used was a Tm:YAG laser (Starmedtec GmbH, Starnberg, Germany) with a wavelength of 1960 nm, a pulse duration of 1 ms and a spot diameter of 5 mm. The physical energy of the painful stimulation was kept constant at 600 mJ. To prevent skin damage, the stimulation site was changed slightly after each stimulus. Tactile stimuli with a force of 181 mN were applied using von Frey monofilaments delivered through a computer-controlled device as described in detail in Dresel et al. (
Interstimulus intervals (ISI) for both modalities were randomly varied between 8 and 12 s. To prevent excessive eye movement related artifacts and blinks the subjects perceived the stimuli with closed eyes. Three seconds after each stimulus, the subjects were prompted by an auditory cue to verbally rate the perceived intensity of the stimulus on a 0–10 numerical rating scale. For pain stimuli this was anchored by no pain (0) and maximum pain (10) the subjects were willing to tolerate during the experiment. For the rating of tactile stimuli, the scale ranged between no perception (0) and maximal imaginable touch (10) that was not perceived as painful.
EEG data were recorded using an electrode cap (FMS, Munich, Germany). The electrode montage included 64 electrodes consisting of all 10–20 system electrodes and the additional electrodes Fpz, FCz, CPz, POz, Oz, Iz, AF3/4, F5/6, FC1/2/3/4/5/6, FT7/8/9/10, C1/2/5/6, CP1/2/3/4/5, TP7/8/9/10, P5/6, PO1/2/9/10, plus two electrodes below the outer canthus of each eye. The EEG was referenced to the FCz electrode, grounded at AFz, sampled at 1 kHz with 0.1 μV resolution. Impedance was kept below 20 kΩ.
Raw EEG data were preprocessed in Vision Analyzer Software (Brain Products, Munich, Germany) including downsampling to 512 Hz, high-pass filtering of 0.5 Hz, correcting for horizontal and vertical eye movements using an independent component analysis, and transforming to the common average reference. Sections of EEG that exceeded ±100 μV in any channel were marked as contaminated with artifacts. Artifact-free trials were epoched from −1100 to 1500 ms and exported to Matlab (The Mathworks, Natick, MA, USA). Time-frequency analyses were performed in Matlab using custom programming on the basis of standard mathematical and signal analysis functions. We applied a single trial Hamming tapered, short-time Fast Fourier Transformation (FFT). The moving window had a length of 100 data points, was padded with zeros up to 512 data points and was shifted by two data points. The frequencies were computed from 2 Hz in steps of 2 Hz (interpolated) up to 100 Hz. On a single trial basis, time-frequency representations (TFRs) were computed and transformed into percent signal change values with respect to the single trial baseline averaged from −1000 to 0 ms. These single trial TFRs were visually inspected for high-frequency artifacts (see examples in “Supplementary Material”). For each subject and electrode, the artifact-free and baseline-corrected single trial TFRs were averaged across trials separately for pain and touch.
In a first step, we determined pain-related and touch-related changes of neuronal activity. For each electrode, group TFRs were calculated by averaging the individual TFRs across subjects. Statistical significant changes of neuronal activity were assessed by calculating paired
In a next step, we related the single trial ratings to the single trial neuronal responses of pain and touch. By using the “lmer” function of the statistical software R
In addition, a further set of LMEs were computed to directly compare the relationships between both modalities, pain and touch (again, across all electrodes but separately for each frequency band). This comparison aimed at elucidating whether pain or touch exhibited a stronger relationship between neuronal responses and perception. Again, the linear function of the model (R command exemplarily shown for theta responses to pain) computes separate parameter values (intercept and slope) for each modality.
For each frequency, the resulting
Laser stimuli elicited moderately painful pinprick-like sensations with a mean subjective pain intensity of 3.7 across subjects. Pain ratings elicited by the repeated application of identical stimuli varied considerably within individuals. A root mean square standard deviation of pain ratings within individuals of 1.5 reveals a substantial intraindividual variability in the perception of pain (Lanier,
For the analysis of EEG data, we first determined neuronal responses to painful and tactile stimuli. TFRs were calculated for each trial and electrode. For the painful stimuli, the group mean TFR at exemplary vertex electrode FCz shows that the brief painful stimuli yielded neuronal responses at latencies between 150 and 1000 ms after stimulus application. We found a strong increase of neuronal activity (210% max. signal change) with a maximum in the theta frequency range (4–8 Hz) at latencies between 150 and 350 ms. The theta response has been shown to be phase-locked to the stimuli and corresponds, hence, to the pain-evoked potential. In addition, theta response and evoked response share the same topographical distribution (Schulz et al.,
For the tactile stimuli, the group mean TFR at exemplary electrode FCz shows similar neuronal responses to those in response to painful stimuli (Figure
To explore and to compare the strength of the relationship between neuronal responses and perception we computed LMEs (Figure
For the tactile domain we revealed that the amplitude of single trials is encoded by neuronal theta responses (all electrodes:
The aim of the present study was to explore, which neurophysiological responses enable the brain to recognize, differentiate and evaluate the intensity of incoming tactile and nociceptive stimuli. We specifically focused on the following two aspects: first, we analyzed the general neuronal responses to both nociceptive and tactile stimuli. Second, we explored how the intensity of both modalities is encoded in the human brain.
For the processing of painful and tactile stimuli, we found a similar pattern of neuronal responses across scalp electrodes, i.e., an increase of gamma and theta activity as well as a decrease of alpha activity. Increased beta activity was found only in the tactile domain.
Increased theta activity was observed in response to both, pain and touch. Amplitudes of tactile theta responses—particularly at central and parietal electrode sites—appear to be smaller than theta responses to pain. The critical question here is whether the difference in amplitude between both modalities can be attributed to a modality-specific processing or to a modality-unspecific difference in saliency. Higher amplitudes of theta oscillations have been shown to reflect the involuntary attention that novel and salient sensory stimuli are drawing (Iannetti et al.,
We further found a stronger decrease of alpha suppression for touch compared to pain as well as a subsequent beta rebound exclusively for tactile responses (Cheyne et al.,
Increased gamma activity was observed in response to both, pain and touch. Amplitudes of tactile gamma responses appear to be smaller than nociceptive gamma responses. Interestingly, touch-related gamma oscillations occurred in a different frequency range (~70 Hz for touch vs. ~80 Hz for pain) and appeared slightly later than pain induced gamma oscillations. This finding suggests that different neuron ensembles—operating in distinct frequencies—contribute to the perception of pain and touch. Previous research has shown separately for pain (Hauck et al.,
The pain-related gamma results of the present experiment are in line with previous other findings (Gross et al.,
In a second step, we related single trial neuronal responses to single trial ratings. To our knowledge, this is the first study that analyzed the relationship between tactile perception and neuro-oscillatory activity in a within-subject design. We found for the tactile domain that low-frequency responses in the theta range encodes for subjective stimulus intensity: theta is positively correlated with touch ratings. Although we revealed a general response of alpha and beta activity in response to all tactile stimuli (see above), these frequencies do not code for the perceived intensity. It seems that alpha/beta activity rather follows an on/off like response characteristic. This could reflect either an involvement in information transmission (or binding), irrespective of different levels of subjective stimulus intensity (Simões et al.,
The analysis of the neuronal coding of laser induced brain signals confirmed recent findings about the role of theta, alpha and gamma activity for the processing of pain (Schulz et al.,
Although the ratings were matched for mean and variance, this does not imply that the rating scales for both modalities are directly comparable. Painful stimuli are commonly believed to be more salient than stimuli of any other sensory modality. This might be particularly true for the present paradigm in which the participants were prompted to keep their eyes closed. Therefore, differences in saliency and attention may cause some of the effects presented here. Neuronal responses in the theta and gamma range have been shown previously to be related to salience (Iannetti et al.,
Despite the application of physically identical stimuli we found a remarkable variability for the perception of pain and touch. Although we would assume that fluctuating cortical processes play an important role for this variability we also need to consider other sources of variability. These sources that were not systematically controlled for include the fiber density of receptors in different skin areas as well as temperature fluctuations of the hand surface. The LME that quantifies the cortical representation of perceptual variability does not distinguish between the cortical and peripheral sources of variability.
Our results demonstrate activity changes in the theta, alpha and gamma range in response to pain. These neuronal responses also encode the intensity of single pain events. For the tactile domain—besides theta, alpha and gamma frequencies—we also revealed increased beta activity in response to all trials. However, the intensity of touch trials was encoded only by theta. Therefore, touch-related alpha and beta responses are suggested to exhibit an on/off like characteristic that is independent from stimulus intensity. The pattern of the present findings, particularly in the alpha, beta and gamma range, suggests that the processing of pain and touch can be attributed to different neuronal ensembles. Further research is needed to investigate the specific contribution of each neuronal oscillation at cellular level, e.g., the mechanisms by which alpha oscillations desynchronize stronger for pain than for touch and encode pain intensity but not touch intensity.
All authors listed, have made substantial, direct and intellectual contribution to the work, and approved it for publication.
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.
We thank Prof Irene Tracey and Dr Meng Liang for their valuable comments on the manuscript. This work was supported by the German Research Foundation (DFG) and the Technische Universität München within the funding programme Open Access Publishing.
The Supplementary Material for this article can be found online at:
1