EEG Recordings as Biomarkers of Pain Perception: Where Do We Stand and Where to Go?

Unraveling the physiological and psychosocial signatures of pain by machine learning

BACKGROUND

Pain is a complex subjective experience, strongly impacting health and quality of life. Despite many attempts to find effective solutions, present treatments are generic, often unsuccessful, and present significant side effects. Designing individualized therapies requires understanding of multidimensional pain experience, considering physical and emotional aspects. Current clinical pain assessments, relying on subjective one-dimensional numeric self-reports, fail to capture this complexity.

METHODS

To this aim, we exploited machine learning to disentangle physiological and psychosocial components shaping the pain experience. Clinical, psychosocial, and physiological data were collected from 118 chronic pain and healthy participants undergoing 40 pain trials (4,697 trials).

FINDINGS

To understand the objective response to nociception, we classified pain from the physiological signals (accuracy >0.87), extracting the most important biomarkers. Then, using multilevel mixed-effects models, we predicted the reported pain, quantifying the mismatch between subjective level and measured physiological response. From these models, we introduced two metrics: TIP (subjective index of pain) and Φ (physiological index). These represent possible added value in the clinical process, capturing psychosocial and physiological pain dimensions, respectively. Patients with high TIP are characterized by frequent sick leave from work and increased clinical depression and anxiety, factors associated with long-term disability and poor recovery, and are indicated for alternative treatments, such as psychological ones. By contrast, patients with high Φ show strong nociceptive pain components and could benefit more from pharmacotherapy.

CONCLUSIONS

TIP and Φ, explaining the multidimensionality of pain, might provide a new tool potentially leading to targeted treatments, thereby reducing the costs of inefficient generic therapies.

FUNDING

RESC-PainSense, SNSF-MOVE-IT197271.

Enhanced EEG power density during painful stretching in individuals with cerebral palsy

BACKGROUND

Pain perception mechanisms in cerebral palsy remain largely unclear.

AIMS

This study investigates brain activity in adults with cerebral palsy during painful and non-painful stretching to elucidate their pain processing characteristics.

METHODS AND PROCEDURES

Twenty adults with cerebral palsy and 20 controls underwent EEG in three conditions: rest, non-painful stretching, and painful stretching. Time-frequency power density of theta, alpha, and beta waves in somatosensory and frontal cortices was analyzed, alongside baseline pressure pain thresholds.

OUTCOMES AND RESULTS

Cerebral palsy individuals exhibited higher theta, alpha, and beta power density in both cortices during painful stretching compared to rest, and lower during non-painful stretching. Controls showed higher power density during non-painful stretching but lower during painful stretching. Cerebral palsy individuals had higher pain sensitivity, with those more sensitive experiencing greater alpha power density.

CONCLUSIONS AND IMPLICATIONS

These findings confirm alterations in the cerebral processing of pain in individuals with cerebral palsy. This knowledge could enhance future approaches to the diagnosis and treatment of pain in this vulnerable population.

Electroencephalographic insights into the pathophysiological mechanisms of emergence delirium in children and corresponding clinical treatment strategies

Emergence delirium is a common postoperative complication in patients undergoing general anesthesia, especially in children. In severe cases, it can cause unnecessary self-harm, affect postoperative recovery, lead to parental dissatisfaction, and increase medical costs. With the widespread use of inhalation anesthetic drugs (such as sevoflurane and desflurane), the incidence of emergence delirium in children is gradually increasing; however, its pathogenesis in children is complex and unclear. Several studies have shown that age, pain, and anesthetic drugs are strongly associated with the occurrence of emergence delirium. Alterations in central neurophysiology are essential intermediate processes in the development of emergence delirium. Compared to adults, the pediatric nervous system is not fully developed; therefore, the pediatric electroencephalogram may vary slightly by age. Moreover, pain and anesthetic drugs can cause changes in the excitability of the central nervous system, resulting in electroencephalographic changes. In this paper, we review the pathogenesis of and prevention strategies for emergence delirium in children from the perspective of brain electrophysiology-especially for commonly used pharmacological treatments-to provide the basis for understanding the development of emergence delirium as well as its prevention and treatment, and to suggest future research direction.

Nociceptive Processing of Elite Athletes Varies between Sport-Specific Loads: An EEG-Based Study Approach

INTRODUCTION

For the downstream nociceptive processing of elite athletes, recent studies indicate that athletes probably tolerate more pain as compared with a normally active population. Phenotyping the nociceptive processing of athletes in different types of endurance sports can provide insight into training-specific effects, which may help in understanding the long-term effects of specific exercise.

METHODS

Twenty-six elite endurance athletes from the disciplines of rowing, triathlon, and running and 26 age- and sex-matched, recreationally active control subjects who participated in the subjective pain perception and processing of standardized noxious stimuli were investigated by EEG. This included standardized heat pain thresholds (HPT) and contact heat-evoked potentials from heat stimulation, measured with EEG as well as pinprick-evoked potentials from mechanical stimulation.

RESULTS

After noxious stimulation, athletes showed a higher activation of the event-related spectral perturbation (ERSP) patterns in the N2P2 EEG response at the Cz Electrode compared with the controls. After noxious contact heat stimulation, triathletes had a higher ERSP activation compared with the controls, whereas the rowers had a higher ERSP activation after noxious mechanical stimulation. Also, HPT in triathletes were increased despite their increased central activation after thermal stimulation. We found a correlation between increased HPT and training hours and years, although athletes did not differ within these variables.

CONCLUSIONS

Although we were able to identify differences between athletes of different endurance sports, the reasons and implications of these differences remain unclear. The study of sport-specific somatosensory profiles may help to understand the mechanisms of exercise-related long-term effects on pain processing and perception. Furthermore, sport-specific somatosensory effects may support the personalization of exercise interventions and identify risk factors for chronic pain in elite athletes.

A Scoping Review of the Effect of EEG Neurofeedback on Pain Complaints in Adults with Chronic Pain

Background and Aim: Non-pharmacological treatments such as electroencephalogram (EEG) neurofeedback have become more important in multidisciplinary approaches to treat chronic pain. The aim of this scoping review is to identify the literature on the effects of EEG neurofeedback in reducing pain complaints in adult chronic-pain patients and to elaborate on the neurophysiological rationale for using specific frequency bands as targets for EEG neurofeedback. Methods: A pre-registered scoping review was set up and reported following the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) extension for Scoping Reviews (PRISMA-ScR). The data were collected by searching for studies published between 1985 and January 2023 in PubMed, EMBASE, and PsycINFO. Results: Thirty-two studies on various types of chronic pain were included. The intervention was well-tolerated. Approximately half of the studies used a protocol that reinforced alpha or sensorimotor rhythms and suppressed theta or beta activity. However, the underlying neurophysiological rationale behind these specific frequency bands remains unclear. Conclusions: There are indications that neurofeedback in patients with chronic pain probably has short-term analgesic effects; however, the long-term effects are less clear. In order to draw more stable conclusions on the effectiveness of neurofeedback in chronic pain, additional research on the neurophysiological mechanisms of targeted frequency bands is definitely worthwhile. Several recommendations for setting up and evaluating the effect of neurofeedback protocols are suggested.

Measurements of pupillary unrest using infrared pupillometry fail to detect changes in pain intensity in patients after surgery: a prospective observational study

PURPOSE

The pupil displays chaotic oscillations, also referred to as pupillary unrest in ambient light (PUAL). As pain has previously been shown to increase pupillary unrest, the quantitative assessment of PUAL has been considered a possible tool to identify and quantify pain. Nevertheless, PUAL is affected by various states, such as vigilance, cognitive load, or emotional arousal, independent of pain. Furthermore, systematically applied opioids are known to reduce PUAL, thus potentially limiting its usefulness to detect pain or changes in pain intensity. To test the hypothesis that PUAL can reliably identify changes in pain intensity in a clinical setting, we measured PUAL in patients experiencing substantial pain relief when regional anesthesia interventions were applied after surgery.

METHODS

We conducted an observational study at an academic surgery centre following institutional review board approval. Eighteen patients with unsatisfactory pain control following surgery underwent regional anesthesia procedures to improve pain control. We used infrared pupillometry to assess pupillary unrest before and after the regional block. We then compared the changes in pupillary unrest with the changes in pain scores (numeric rating scale [NRS], range 0-10).

RESULTS

Eighteen patients received epidural anesthesia (n = 14) or peripheral nerve blocks (n = 4), resulting in improvement of mean (standard deviation [SD]) NRS pain scores from 7.2 (1.7) to 1.9 (1.8) (difference in means, -2.2; 95% confidence interval [CI], -6.3 to -4.1; P < 0.001). Nevertheless, pupillary unrest did not change as pain decreased; the mean (SD) PUAL was 0.113 (0.062) before analgesia and 0.112 (0.068) after analgesia (difference in means, -0.001; 95% CI, -0.018 to 0.015; P = 0.88).

CONCLUSION

In this prospective observational study, pupillometric measurements of pupillary unrest did not identify changes in pain intensity in a postoperative, predominantly opioid-exposed patient population. While the sample size was small, the use of measurements of pupillary unrest to detect and quantify pain has to be questioned.

Resting-state electroencephalography delta and theta bands as compensatory oscillations in chronic neuropathic pain: a secondary data analysis

Chronic neuropathic pain (CNP) remains a significant clinical challenge, with complex neurophysiological underpinnings that are not fully understood. Identifying specific neural oscillatory patterns related to pain perception and interference can enhance our understanding and management of CNP. To analyze resting electroencephalography data from individuals with chronic neuropathic pain to explore the possible neural signatures associated with pain intensity, pain interference, and specific neuropathic pain characteristics. We conducted a secondary analysis from a cross-sectional study using electroencephalography data from a previous study, and Brief Pain Inventory from 36 patients with chronic neuropathic pain. For statistical analysis, we modeled a linear or logistic regression by dependent variable for each model. As independent variables, we used electroencephalography data with such brain oscillations: as delta, theta, alpha, and beta, as well as the oscillations low alpha, high alpha, low beta, and high beta, for the central, frontal, and parietal regions. All models tested for confounding factors such as age and medication. There were no significant models for Pain interference in general activity, walking, work, relationships, sleep, and enjoyment of life. However, the model for pain intensity during the past four weeks showed decreased alpha oscillations, and increased delta and theta oscillations were associated with decreased levels of pain, especially in the central area. In terms of pain interference in mood, the model showed high oscillatory Alpha signals in the frontal and central regions correlated with mood impairment due to pain. Our models confirm recent findings proposing that lower oscillatory frequencies, likely related to subcortical pain sources, may be associated with brain compensatory mechanisms and thus may be associated with decreased pain levels. On the other hand, higher frequencies, including alpha oscillations, may disrupt top-down compensatory mechanisms.

A novel theta-controlled vibrotactile brain-computer interface to treat chronic pain: a pilot study

Limitations in chronic pain therapies necessitate novel interventions that are effective, accessible, and safe. Brain-computer interfaces (BCIs) provide a promising modality for targeting neuropathology underlying chronic pain by converting recorded neural activity into perceivable outputs. Recent evidence suggests that increased frontal theta power (4-7 Hz) reflects pain relief from chronic and acute pain. Further studies have suggested that vibrotactile stimulation decreases pain intensity in experimental and clinical models. This longitudinal, non-randomized, open-label pilot study's objective was to reinforce frontal theta activity in six patients with chronic upper extremity pain using a novel vibrotactile neurofeedback BCI system. Patients increased their BCI performance, reflecting thought-driven control of neurofeedback, and showed a significant decrease in pain severity (1.29 ± 0.25 MAD, p = 0.03, q = 0.05) and pain interference (1.79 ± 1.10 MAD p = 0.03, q = 0.05) scores without any adverse events. Pain relief significantly correlated with frontal theta modulation. These findings highlight the potential of BCI-mediated cortico-sensory coupling of frontal theta with vibrotactile stimulation for alleviating chronic pain.

The temporal effect of uncertain context on the perceptual processing of painful and non-painful stimulation

Uncertainty has been demonstrated to influence the perception of noxious stimuli, but little is known about the effects of prolonged uncertain contexts on the perception of painful and non-painful stimuli. To address this knowledge gap, the present study utilized a cue-based NPU-threat task, where uncertain and certain trials were separated into distinct blocks. The objective was to investigate the impact of uncertain contexts on the temporal dynamics of electroencephalogram (EEG) activity during the processing of painful and non-painful stimuli. The results revealed that the influence of uncertain contexts on neural responses extends beyond painful trials and is also evident in non-painful trials. In uncertain contexts, it has been observed that painful stimuli elicit larger P2 amplitudes and late beta band (13-30 Hz) event-related desynchronization (ERD) around 500-700 ms. However, in certain contexts, painful stimuli evoke stronger late gamma band (50-70 Hz) event-related synchronization (ERS) around 600-700 ms. For non-painful trials, in uncertain contexts, significantly higher amplitudes of the late positive potential (LPP) component and delta-theta band (2-7 Hz) ERS were observed compared to certain non-painful stimuli. These findings demonstrate that uncertain contexts exert a significant impact on the processing of both painful and non-painful stimuli, and this influence is mediated by distinct neural mechanisms.

Correlation between spinal cord stimulation analgesia and cortical dynamics in pain management

AIM

Spinal cord stimulation (SCS) is an effective method to treat neuropathic pain. It is necessary to identify the responders of SCS analgesia before implantation. The aim of this study is to investigate the relationship between the cortical dynamics and SCS analgesia responders in pain management.

METHODS

Resting-state EEG recording was performed in patients who underwent short-term implantation of spinal cord stimulation for pain therapy. We then did spectral analysis to capture the pattern of cortical oscillation between neuromodulation therapy analgesia responders and nonresponders.

RESULTS

About 58.3% (14 out of 24) of participants were considered as analgesia responders, with average visual analogue scores reduction of 4.8 ± 1.0 after surgery, and 2.1 ± 0.7 for the nonresponder subgroup, respectively. The alpha oscillation was significantly enhanced in responder cohort compared with nonresponders. We also observed an increasing spectral power of gamma band in responders. Furthermore, the attenuation of pain severity was significantly correlated with the global alpha oscillation activity (r = 0.60, P = 0.002). Likely, positive and significant correlation was found between the pain relief and gamma activity (r = 0.58, P = 0.003).

CONCLUSIONS

Distinct pattern of neural oscillation is associated with the analgesic effect of spinal cord stimulation in pain management, enhancement of cortical alpha and gamma oscillation may be a predictor of analgesia responders.

Electroencephalography-Based Effects of Acute Alcohol Intake on the Pain Matrix

The effects of acute and chronic intakes of high doses of alcohol on pain perception are well known, ranging from short-term analgesic effects to long-term sensitization and polyneuropathies. The short-term analgesic effects of ethanol consumption on subjective pain perception have been well studied in the literature. Recent advances in neuroimaging allow for an insight into pain-related structures in the brain, fostering the mechanistic understanding of the processing of nociceptive input and pain. We aimed to utilize EEG, combined with standardized noxious mechanical/thermal stimulation and subjective pain testing, to research the effects of acute alcohol intake on nociceptive processing and pain perception. We recruited 12 healthy subjects in an unblinded cross-over study design and aimed at achieving a blood alcohol level of 0.1%. Our data revealed a significant reduction in subjective pain ratings to noxious thermal and mechanical stimuli after alcohol ingestion. Our EEG data revealed suppressing effects on the cortical structures responsible for processing pain, the "pain matrix". We conclude that in addition to its analgesic effects, as expressed by the reduction in subjective pain, alcohol has a further impact on the "pain matrix" and directly affects the salience to a nociceptive stimulus.

Brain-Computer Interface to Deliver Individualized Multisensory Intervention for Neuropathic Pain

To unravel the complexity of the neuropathic pain experience, researchers have tried to identify reliable pain signatures (biomarkers) using electroencephalography (EEG) and skin conductance (SC). Nevertheless, their use as a clinical aid to design personalized therapies remains scarce and patients are prescribed with common and inefficient painkillers. To address this need, novel non-pharmacological interventions, such as transcutaneous electrical nerve stimulation (TENS) to activate peripheral pain relief via neuromodulation and virtual reality (VR) to modulate patients' attention, have emerged. However, all present treatments suffer from the inherent bias of the patient's self-reported pain intensity, depending on their predisposition and tolerance, together with unspecific, pre-defined scheduling of sessions which does not consider the timing of pain episodes onset. Here, we show a Brain-Computer Interface (BCI) detecting in real-time neurophysiological signatures of neuropathic pain from EEG combined with SC and accordingly triggering a multisensory intervention combining TENS and VR. After validating that the multisensory intervention effectively decreased experimentally induced pain, the BCI was tested with thirteen healthy subjects by electrically inducing pain and showed 82% recall in decoding pain in real time. Such constructed BCI was then validated with eight neuropathic patients reaching 75% online pain precision, and consequently releasing the intervention inducing a significant decrease (50% NPSI score) in neuropathic patients' pain perception. Our results demonstrate the feasibility of real-time pain detection from objective neurophysiological signals, and the effectiveness of a triggered combination of VR and TENS to decrease neuropathic pain. This paves the way towards personalized, data-driven pain therapies using fully portable technologies.

In search of a composite biomarker for chronic pain by way of EEG and machine learning: where do we currently stand?

Machine learning is becoming an increasingly common component of routine data analyses in clinical research. The past decade in pain research has witnessed great advances in human neuroimaging and machine learning. With each finding, the pain research community takes one step closer to uncovering fundamental mechanisms underlying chronic pain and at the same time proposing neurophysiological biomarkers. However, it remains challenging to fully understand chronic pain due to its multidimensional representations within the brain. By utilizing cost-effective and non-invasive imaging techniques such as electroencephalography (EEG) and analyzing the resulting data with advanced analytic methods, we have the opportunity to better understand and identify specific neural mechanisms associated with the processing and perception of chronic pain. This narrative literature review summarizes studies from the last decade describing the utility of EEG as a potential biomarker for chronic pain by synergizing clinical and computational perspectives.

Analysis two types of K complexes on the human EEG based on classical continuous wavelet transform

In our work, we compare EEG time-frequency features for two types of K-complexes detected in volunteers performing the monotonous psychomotor test with their eyes closed. Type I K-complexes preceded spontaneous awakenings, while after type II K-complexes, subjects continued to sleep at least for 10 s after. The total number of K-complexes in the group of 18 volunteers was 646, of which of which type I K-complexes was 150 and type II K-complexes was 496. Time-frequency analysis was performed using continuous wavelet transform. EEG wavelet spectral power was averaged upon several brain zones for each of the classical frequency ranges (slow wave, δ, θ, α, β₁, β₂, γ bands). The low-frequency oscillatory activity ( δ-band) preceding type I K-complexes was asymmetrical and most prominent in the left hemisphere. Statistically significant differences were obtained by averaging over the left and right hemispheres, as well as projections of the motor area of the brain, p<0.05. The maximal differences between the types I and II of K-complexes were demonstrated in δ-, θ-bands in the occipital and posterior temporal regions. The high amplitude of the motor cortex projection response in β₂-band, [20;30] Hz, related to the sensory-motor modality of task in monotonous psychomotor test. The δ-oscillatory activity preceding type I K-complexes was asymmetrical and most prominent in the left hemisphere may be due to the important role of the left hemisphere in spontaneous awakening from sleep during monotonous work, which is an interesting issue for future research.

EEG responses to standardised noxious stimulation during clinical anaesthesia: a pilot study

Background

During clinical anaesthesia, the administration of analgesics mostly relies on empirical knowledge and observation of the patient's reactions to noxious stimuli. Previous studies in healthy volunteers under controlled conditions revealed EEG activity in response to standardised nociceptive stimuli even at high doses of remifentanil and propofol. This pilot study aims to investigate the feasibility of using these standardised nociceptive stimuli in routine clinical practice.

Methods

We studied 17 patients undergoing orthopaedic trauma surgery under general anaesthesia. We evaluated if the EEG could track standardised noxious phase-locked electrical stimulation and tetanic stimulation, a time-locked surrogate for incisional pain, before, during, and after the induction of general anaesthesia. Subsequently, we analysed the effect of tetanic stimulation on the surgical pleth index as a peripheral, vegetative, nociceptive marker.

Results

We found that the phase-locked evoked potentials after noxious electrical stimulation vanished after the administration of propofol, but not at low concentrations of remifentanil. After noxious tetanic stimulation under general anaesthesia, there were no consistent spectral changes in the EEG, but the vegetative response in the surgical pleth index was statistically significant (Hedges' g effect size 0.32 [95% confidence interval 0.12-0.77], P=0.035).

Conclusion

Our standardised nociceptive stimuli are not optimised for obtaining consistent EEG responses in patients during clinical anaesthesia. To validate and sufficiently reproduce EEG-based standardised stimulation as a marker for nociception in clinical anaesthesia, other pain models or stimulation settings might be required to transfer preclinical studies into clinical practice.

Clinical trial registration

DRKS00017829.

Measuring pain and nociception: Through the glasses of a computational scientist. Transdisciplinary overview of methods

In a healthy state, pain plays an important role in natural biofeedback loops and helps to detect and prevent potentially harmful stimuli and situations. However, pain can become chronic and as such a pathological condition, losing its informative and adaptive function. Efficient pain treatment remains a largely unmet clinical need. One promising route to improve the characterization of pain, and with that the potential for more effective pain therapies, is the integration of different data modalities through cutting edge computational methods. Using these methods, multiscale, complex, and network models of pain signaling can be created and utilized for the benefit of patients. Such models require collaborative work of experts from different research domains such as medicine, biology, physiology, psychology as well as mathematics and data science. Efficient work of collaborative teams requires developing of a common language and common level of understanding as a prerequisite. One of ways to meet this need is to provide easy to comprehend overviews of certain topics within the pain research domain. Here, we propose such an overview on the topic of pain assessment in humans for computational researchers. Quantifications related to pain are necessary for building computational models. However, as defined by the International Association of the Study of Pain (IASP), pain is a sensory and emotional experience and thus, it cannot be measured and quantified objectively. This results in a need for clear distinctions between nociception, pain and correlates of pain. Therefore, here we review methods to assess pain as a percept and nociception as a biological basis for this percept in humans, with the goal of creating a roadmap of modelling options.

Whether ambulatory electroencephalogram and visual tracking system could be the new strategy for pain assessment?

The human pain experience is a complex multi-faceted symptom. Effective pain management begins with a comprehensive assessment. However, a plethora of existing assessment tools for pain assessment focus more on self-report of pain intensity but lack of multi-dimensional impersonal assessment. These unidimensional scales, which capture self-reported levels of pain intensity, not only underestimate the complexity of the pain experience, but also lack stability and objectivity in their own assessments of pain intensity. Therefore, we propose a hypothesis that using scientific and technological means, such as visual tracking and surveillance system, ambulatory electroencephalogram and other techniques, combined with psychological assessment pictures and existing scales, to comprehensively evaluate pain may provide a new method for more effective clinical treatment of pain, especially chronic severe pain.

External validation of binary machine learning models for pain intensity perception classification from EEG in healthy individuals

Discrimination of pain intensity using machine learning (ML) and electroencephalography (EEG) has significant potential for clinical applications, especially in scenarios where self-report is unsuitable. However, existing research is limited due to a lack of external validation (assessing performance using novel data). We aimed for the first external validation study for pain intensity classification with EEG. Pneumatic pressure stimuli were delivered to the fingernail bed at high and low pain intensities during two independent EEG experiments with healthy participants. Study one (n = 25) was utilised for training and cross-validation. Study two (n = 15) was used for external validation one (identical stimulation parameters to study one) and external validation two (new stimulation parameters). Time-frequency features of peri-stimulus EEG were computed on a single-trial basis for all electrodes. ML training and analysis were performed on a subset of features, identified through feature selection, which were distributed across scalp electrodes and included frontal, central, and parietal regions. Results demonstrated that ML models outperformed chance. The Random Forest (RF) achieved the greatest accuracies of 73.18, 68.32 and 60.42% for cross-validation, external validation one and two, respectively. Importantly, this research is the first to externally validate ML and EEG for the classification of intensity during experimental pain, demonstrating promising performance which generalises to novel samples and paradigms. These findings offer the most rigorous estimates of ML's clinical potential for pain classification.