Abstract The somatosensory system encodes peripheral inputs through a sequence of ascending neural relays spanning spinal, subcortical and cortical levels. While multivariate decoding of electroencephalography (EEG) signals has demonstrated that cortical activity contains fine-grained information about somatosensory stimuli, the extent to which earlier processing stages contribute additional, non-redundant information remains unclear. To address this gap, we investigated a dataset comprising peripheral sensory stimulation and mixed stimulation (i.e., stimulation engaging both sensory and motor fibers). We assessed whether stimulation characteristics can be decoded from spinal recordings using high-density electrospinography (ESG), and whether combining ESG with EEG enhances decoding performance. Decoding accuracy varied systematically with both stimulation type and signal modality. ESG was the most informative signal for mixed and mixed vs sensory discrimination, reaching an average accuracy of ∼98%, while EEG provided a relative advantage for purely sensory tasks, though absolute accuracy remained more modest for both modalities. Critically, combining the two modalities together consistently matched or outperformed either one, used alone, across all conditions, with gains most pronounced for mixed vs sensory discrimination. Multi-subject generalization improved progressively with training-set size, rising to ∼88% with 15 training subjects for mixed classification, suggesting that subject-independent decoding of motor intent may be achievable when models are trained on a larger number of subjects. Taken together, these results establish that spinal ESG signals carry decodable information about peripheral stimulation that is complementary to and not redundant with cortical EEG. This finding supports a multilevel framework for decoding sensorimotor processing in humans and motivates the development of dual-modality brain–machine interfaces that leverage both cortical and spinal signals to improve the control of neurostimulation and assistive devices.

The spinal cord is an important component of the central nervous system for the processing of sensorimotor information transmitted between the body and the brain. Electrospinography (ESG) is the most accessible non-invasive technique for recording spinal signals in humans, but the detrimental impact of physiological noise (mostly of cardiac nature) has prevented widespread adoption. Here, we aim to address this issue by examining various denoising algorithms for cardiac artefact reduction-including approaches based on principal component analysis (PCA), independent component analysis (ICA), signal space projection (SSP), canonical correlation analysis (CCA), and denoising separation of sources (DSS). We observed that in situations where large number of spinal electrodes are used, ICA and SSP offer the best results in terms of balancing the removal of noise and preserving neural information of interest. In cases where only a small number of electrodes are available, an approach based on PCA is deemed helpful. Finally, we also approached this issue from a signal-enhancement perspective by applying CCA and DSS directly to signals of interest, namely spinal somatosensory evoked potentials (SEPs, especially the N13 and N22 components in the cervical and lumbar spinal cord, respectively). We observed that in cases where extensive electrode arrays are used in the context of task-based designs, CCA reveals clear evoked spinal potentials even with single-trial resolution. Taken together, there are several appropriate algorithms for physiological noise removal and/or signal enhancement in ESG, rendering this an accessible and easy-to-use technique for non-invasive assessments of human spinal cord function.

Repetition suppression, the reduced neural response upon repeated presentation of a stimulus, can be explained by models focusing on bottom-up (i.e., adaptation) or top-down (i.e., expectation) mechanisms. Predictive coding models fall into the latter category and propose that repetitions are expected and therefore elicit smaller prediction error responses. Although studies in the visual and auditory domain provide some support for such models, in nociception evidence remains inconclusive, despite the substantial influence expectations exert on pain perception. To assess expectation effects on repetition suppression in nociception, we developed a paradigm in which healthy volunteers received brief CO2 laser stimuli, while we acquired electroencephalographic (EEG) and peripheral physiological data. Importantly, laser stimuli could be either repeated after one second or not be repeated, with the probability of repetitions manipulated in a block-wise fashion, such that repetitions were either expected or unexpected. We observed repetition suppression in laser-evoked potentials and laser-induced gamma band oscillations, but not in laser-induced desynchronizations in the alpha and beta band. Critically, neither these EEG responses, nor the peripheral physiological data showed significant differences between the expectation conditions, with Bayesian analyses mostly providing evidence for an absence of effects. This indicates that repetition suppression to brief nociceptive laser stimuli is not driven by top-down factors, but rather mediated by other adaptation processes. Although this does not preclude an influence of predictive coding models in nociception, it suggests that when the nervous system receives highly precise input, its responses are less susceptible to influence from expectations.NEW & NOTEWORTHY We show that repetition suppression (RS; the diminished neural activity upon repeated stimulus presentation) can be observed in the nociceptive domain: brief laser stimuli lead to RS in event-related potentials and gamma oscillations. Importantly, nociceptive RS is not modulated by repetition probability and thus expectations regarding the occurrence of a repetition. This indicates that nociceptive responses are less prone to top-down expectations under conditions of limited sensory uncertainty, as established via precise laser stimuli.

The spinal cord is of fundamental importance for integrative processing in brain-body communication, yet routine noninvasive recordings in humans are hindered by vast methodological challenges. Here, we overcome these challenges by developing an easy-to-use electrophysiological approach based on high-density multichannel spinal recordings combined with multivariate spatial-filtering analyses. These advances enable a spatiotemporal characterization of spinal cord responses and demonstrate a sensitivity that permits assessing even single-trial responses. To furthermore enable the study of integrative processing along the neural processing hierarchy in somatosensation, we expand this approach by simultaneous peripheral, spinal, and cortical recordings and provide direct evidence that bottom-up integrative processing occurs already within the spinal cord and thus after the first synaptic relay in the central nervous system. Finally, we demonstrate the versatility of this approach by providing noninvasive recordings of nociceptive spinal cord responses during heat-pain stimulation. Beyond establishing a new window on human spinal cord function at millisecond timescale, this work provides the foundation to study brain-body communication in its entirety in health and disease.

Sensory information processing in the central nervous system (CNS) requires both post-synaptic responses and spiking activity, with human electroencephalography (EEG) research largely focused on the former. However, EEG data also contain high-frequency oscillations (HFOs), which have been linked to cortical spike bursts in animal models, thus providing a non-invasive macroscopic marker of population spiking in humans. Here, we go beyond cortical processing and demonstrate that it is possible to simultaneously record HFOs to somatosensory stimulation across the entire human CNS - from spinal cord over subcortical to cortical processing stages. Using multivariate spatial filtering approaches in two independent datasets, we identify replicable and generalizable HFOs across the CNS and obtain a detailed characterization of these signals at each processing level. Finally, we provide evidence at the between- and within-individual level that low and high frequency electrophysiological responses represent at least partly independent information across the CNS. Taken together, our approach offers the first evidence that it is possible to detect and characterise HFOs simultaneously across the CNS, providing a unique non-invasive and multi-level window into human neurophysiology in health and disease. ### Competing Interest Statement The authors have declared no competing interest.

Identical sensory stimuli can lead to different neural responses depending on the instantaneous brain state. Specifically, neural excitability in sensory areas may shape the brain´s response already from earliest cortical processing onwards. However, whether these dynamics affect a given sensory domain as a whole or occur on a spatially local level is largely unknown. We studied this in the somatosensory domain of 38 human participants with EEG, presenting stimuli to the median and tibial nerves alternatingly, and testing the co-variation of initial cortical responses in hand and foot areas, as well as their relation to pre-stimulus oscillatory states. We found that amplitude fluctuations of initial cortical responses to hand and foot stimulation – the N20 and P40 components of the somatosensory evoked potential (SEP), respectively – were not related, indicating local excitability changes in primary sensory regions. In addition, effects of pre-stimulus alpha (8-13 Hz) and beta (18-23 Hz) band amplitude on hand-related responses showed a robust somatotopic organization, thus further strengthening the notion of local excitability fluctuations. However, for foot-related responses, the spatial specificity of pre-stimulus effects was less consistent across frequency bands, with beta appearing to be more foot-specific than alpha. Connectivity analyses in source space suggested this to be due to a somatosensory alpha rhythm that is primarily driven by activity in hand regions while beta frequencies may operate in a more hand-region-independent manner. Altogether, our findings suggest spatially distinct excitability dynamics within the primary somatosensory cortex, yet with the caveat that frequency-specific processes in one sub-region may not readily generalize to other sub-regions.

The spinal cord is of fundamental importance for somatosensory processing and plays a significant role in various pathologies, such as chronic pain. However, knowledge on spinal cord processing in humans is limited due to the vast technical challenges involved in its investigation via non-invasive recording approaches. Here, we aim to address these challenges by developing an electrophysiological approach – based on a high-density electrode-montage – that allows for characterizing spinal cord somatosensory evoked potentials (SEPs) and combining this with concurrent recordings of the spinal cord’s input (peripheral nerve action potentials) and output (SEPs in brainstem and cortex). In two separate experiments, we first methodologically validate the approach (including replication and robustness analyses) and then assess its application in the context of a neuroscientific question (integrative processes along the neural hierarchy). Critically, we demonstrate the benefits of multi-channel recordings in terms of enhancing sensitivity via spatial filtering, which also allows for obtaining spinal cord SEPs at the single-trial level. We make use of this approach to demonstrate the feasibility of recording spinal cord SEPs in low-signal scenarios (single-digit stimulation) and – most importantly – to provide evidence for bottom-up signal integration already at the level of the spinal cord. Taken together, our approach of concurrent multi-channel recordings of evoked responses along the neural hierarchy allows for a comprehensive assessment of the functional architecture of somatosensory processing at a millisecond timescale.### Competing Interest StatementThe authors have declared no competing interest.

Brain-computer interfacing (BCI) that reads brain activity and generates commands to control the movements of the body—real, virtual, prosthetic or robotic—can be used as a means for neurorehabilitation. BCI can be used to carry out motor actions which, being lost through the real body, bypass it and use other effectors. However, do BCI-generated body movements result in feelings of agency? Our findings highlight the advantages of motor-imagery-based BCI protocols over steady-state visual-evoked potentials for their use in controlling surrogate bodies. BCI paradigms that use the activation of sensorimotor areas (motor imagery) induce greater agency and sense of responsibility, resulting in a more effective user-experience for neurorehabilitation.

Changes in body representation may affect pain perception. The effect of a distorted body image, such as the telescoping effect in amputee patients, on pain perception, is unclear. This study aimed to investigate whether distorting an embodied virtual arm in virtual reality (simulating the telescoping effect in amputees) modulated pain perception and anticipatory responses to pain in healthy participants. Twenty-seven right-handed participants were immersed in virtual reality and the virtual arm was shown with three different levels of distortion with a virtual threatening stimulus either approaching or contacting the virtual hand. We evaluated pain/discomfort ratings, ownership, and skin conductance responses (SCRs) after each condition. Viewing a distorted virtual arm enhances the SCR to a threatening event with respect to viewing a normal control arm, but when viewing a reddened-distorted virtual arm, SCR was comparatively reduced in response to the threat. There was a positive relationship between the level of ownership over the distorted and reddened-distorted virtual arms with the level of pain/discomfort, but not in the normal control arm. Contact with the threatening stimulus significantly enhances SCR and pain/discomfort, while reduced SCR and pain/discomfort were seen in the simulated-contact condition. These results provide further evidence of a bi-directional link between body image and pain perception.

Agency is the attribution of an action to the self and is a prerequisite for experiencing responsibility over its consequences. Here we investigated agency and responsibility by studying the control of movements of an embodied avatar, via brain computer interface (BCI) technology, in immersive virtual reality. After induction of virtual body ownership by visuomotor correlations, healthy participants performed a motor task with their virtual body. We compared the passive observation of the subject’s ‘own’ virtual arm performing the task with (1) the control of the movement through activation of sensorimotor areas (motor imagery) and (2) the control of the movement through activation of visual areas (steady-state visually evoked potentials). The latter two conditions were carried out using a brain–computer interface (BCI) and both shared the intention and the resulting action. We found that BCI-control of movements engenders the sense of agency, which is strongest for sensorimotor areas activation. Furthermore, increased activity of sensorimotor areas, as measured using EEG, correlates with levels of agency and responsibility. We discuss the implications of these results for the neural bases of agency, but also in the context of novel therapies involving BCI and the ethics of neurotechnology. Key points summary We induced embodiment of a virtual body and its movements were controlled by two different BCI paradigms – one based on signals from sensorimotor versus one from visual cortical areas. BCI-control of movements engenders agency, but not equally for all paradigms. Cortical sensorimotor activation correlates with agency and responsibility. This has significant implications for neurological rehabilitation and neuroethics.

The most robust and clear biological index differentiating persons with schizophrenia from healthy controls is the drastic reduction of the amplitude of their P300b event-related brain potential (ERP). However, the cause of that reduction remains obscure. Nevertheless, the P300b belongs to the family of the late posterior positivities (LPPs) which are closely related to the consciousness of the meaning of the stimulus in the task for the participants themselves (e.g., the: I am seeing the target stimulus for which I have to respond). The fragmentation of the self present in schizophrenia, could thus be the cause. If this were true, then P300bs should be somewhat reduced in healthy participants when their self representations are temporarily and minimally fragmented. We tested this hypothesis by using the innocuous fragmentation of the self that occurs in virtual reality (VR). There, participants can have a fragment of their self in an avatar they feel embodied in, within a VR room, while having another fragment of their self in their real body in the real room where they know they are. Our participants were thus equipped with a head mounted display in which they viewed a virtual room where a female humanoid avatar was facing them. She was lifting her right hand in synchrony with the participants, in order to induce in them a feeling of embodiment. Stimuli were a frequent green- and a rare red-disk, the oddball stimulus, occurring over the right hand of the avatar. Participants had to perform a Go/NoGo task, lifting their right hand to the frequent green disk and repressing this action for the oddball red disk. In the syncMove block of trials the avatar was lifting her right hand synchronously with the participant, disturbing her self representation as confirmed by the debriefing session. In the noMove block, the avatar remained immobile. In the classic block, only the red and the green disk were displayed on a monochrome background, neither the room nor the avatar were shown. As predicted, P300bs were found to be smaller in the syncMove block than in the noMove- and the classic-block in participants who had the classically large P300b oddball effect between ERPs to the frequent and those to the rare stimuli. Reduced P300bs of schizophrenia could thus be partly due to self fragmentation. Results may also open an avenue of research to the functional significance of LPPs and the content of the consciousness indexed by these potentials.

La experiencia de saber que este cuerpo es mi cuerpo se entiende por body ownership y esta estrechamente relacionada con la conciencia. Se cree que el body ownership emerge de procesos de integracion multisensorial, tal y como se ha demostrado a traves de la ilusion de la mano de goma. Esta ilusion induce la sensacion de ownership sobre una mano de goma tras ocultar la mano real de la persona fuera de su vista y estimulandolas de forma simultanea con un pincel. El resultado es la experiencia ilusoria de que la mano de goma es parte de nuestro propio cuerpo. Las ilusiones de body ownership tambien se pueden experimentar en realidad virtual (RV) inmersiva, metodo que se ha utilizado en los tres experimentos llevados a cabo en esta tesis.\par El primer experimento de esta tesis tenia el objetivo de investigar los mecanismos subyacentes en el body ownership. Los experimentos sobre la percepcion tactil, visual y auditiva mediante registros electroencefalograficos muestran que la actividad neuronal espontanea, que es una actividad neuronal relacionada con inputs u outputs especificos del cerebro, puede predecir la percepcion de un estimulo. La actividad neuronal espontanea medida con fMRI tipicamente fluctua entre 0,01 y 0,1Hz y se cree que esta organizada en redes, tambien denominadas redes en estado de reposo. Estas redes se han relacionado con funciones cognitivas, somatosensoriales y sensoriales. Algunas personas que participaron en experimentos de body ownership en nuestro laboratorio reportaron sentir cambios en el body ownership a lo largo del tiempo, planteando la pregunta de si tales fluctuaciones de la ilusion del body ownership estan presentes en mas personas y si es posible predecir estas fluctuaciones de la ilusion con la actividad espontanea del cerebro. Para investigar esto, medimos los cambios en la ilusion de body ownership con una tarea de simultaneidad visuo-tactil en dos condiciones diferentes. En una condicion el participante experimento un alto grado de body ownership y en la otra un bajo grado. Encontramos diferencias en el umbral de simultaneidad requerido para la tarea de simultaneidad visuo-tactil entre las condiciones de alto y bajo body ownership. Tambien encontramos que la experimentacion de un alto y bajo nivel de body ownership fluctua en el tiempo tiempo, lo que podria reflejar fluctuaciones en la integracion multisensorial. Mas investigacion es requerida para comprender la relacion entre la actividad cerebral espontanea y las fluctuaciones en el body ownership. Los resultados se discuten en relacion al diseno experimental utilizado. El segundo experimento tenia el objetivo de investigar la relacion entre las ilusiones de body ownership y la percepcion del dolor. Se ha demostrado que la observacion del propio cuerpo puede tener propiedades analgesicas. Este efecto ampliamente conocido en relacion a nuestra mano real ha sido tambien estudiado en manos poseidas de forma ilusoria encontrando resultados contradictorios. Mientras que el efecto analgesico de mirar a la propia mano se ha replicado en embodiment inducido en RV, existen hallazgos polemicos en referencia a la ilusion de la mano de goma: algunos autores encuentran un efecto analgesico al mirar la mano de goma poseida y otros no. Una diferencia crucial entre la ilusion de la mano de goma y el embodiment inducido en RV es que en RV se puede colocar la mano real y virtual en la misma localizacion, mientras que por obvias razones esto no es posible en la ilusion de la mano de goma. Para probar si la distancia entre la mano real y la mano artificial poseida puede explicar los hallazgos contradictorios en la literatura hemos manipulado la distancia y la estimulacion visuo-tactil en un diseno experimental bifactorial midiendo los umbrales de dolor individual (HPT) de los participantes. Nuestros datos mostraron que el HPT era mayor cuando ambas manos se colocaban en la misma localizacion en comparacion a cuando estaban situadas a 30 cm de distancia, mientras que la simultaneidad de la estimulacion visuo-tactil no influyo en el HPT. Tambien encontramos una relacion positiva entre los niveles reportados de body ownership y HPT. Los resultados obtenidos se discuten en referencia a la bibliografia reciente y con respecto a sus implicaciones terapeuticas. Cuando las personas experimentan altos niveles de body ownership sobre un cuerpo virtual que realiza acciones, esas acciones pueden atribuirse erroneamente al yo. La auto-atribucion de una accion se denomina agency y nos da la sensacion de control y responsabilidad sobre nuestras propias acciones. La agency ha sido descrita como resultado de una correspondencia entre la retroalimentacion sensorial predicha y la real en una accion motora planificada, un proceso que involucra areas motoras. Sin embargo, situaciones en que la experimentacion de body ownership resulta tambien en una ilusion de agency sobre un cuerpo artificial plantean la cuestion de si tambien existe la implicacion de areas motoras en el sentido de agency. En el tercer experimento de esta tesis exploramos esta cuestion en el contexto de las interfaces cerebro-ordenador (BCI). En dos condiciones, los participantes controlaron el movimiento de un brazo virtual a traves de una BCI que utilizaba la actividad de areas motoras o visuales contra una tercera condicion control en que los participantes simplemente observaron el movimiento. Nuestros datos muestran que las puntuaciones mas altas en el sentido de agency fueron obtenidas al utilizar areas motoras, mientras que las puntuaciones con areas visuales fueron las segundas mas altas y las puntuaciones mas bajas sucedieron en la observacion del movimiento virtual. Curiosamente, la responsabilidad sobre las consecuencias de la accion solo fue inducida cuando el movimiento virtual era controlado a traves de las areas motoras. Adicionalmente, mientras mas alta era la actividad en las areas motoras, mas altos fueron las puntuaciones reportadas en el sentido de responsabilidad. Se discute la contribucion que estos resultados suponen para nuestro conocimiento del sentido de agency en base a como diferentes protocolos de BCI afectan el sentido de agency y en sus implicaciones terapeuticas. En conjunto, estos experimentos investigaron los procesos subyacentes en el body ownership y sus influencias en la percepcion del dolor y en el sentido de agency. Los resultados tienen implicaciones en la gestion del dolor y la rehabilitacion neurologica.

Seeing one's own body has been reported to have analgesic properties. Analgesia has also been described when seeing an embodied virtual body colocated with the real one. However, there is controversy regarding whether this effect holds true when seeing an illusory-owned body part, such as during the rubber-hand illusion. A critical difference between these paradigms is the distance between the real and surrogate body part. Colocation of the real and surrogate arm is possible in an immersive virtual environment, but not during illusory ownership of a rubber arm. The present study aimed at testing whether the distance between a real and a virtual arm can explain such differences in terms of pain modulation. Using a paradigm of embodiment of a virtual body allowed us to evaluate heat pain thresholds at colocation and at a 30-cm distance between the real and the virtual arm. We observed a significantly higher heat pain threshold at colocation than at a 30-cm distance. The analgesic effects of seeing a virtual colocated arm were eliminated by increasing the distance between the real and the virtual arm, which explains why seeing an illusorily owned rubber arm does not consistently result in analgesia. These findings are relevant for the use of virtual reality in pain management.PERSPECTIVE:Looking at a virtual body has analgesic properties similar to looking at one's real body. We identify the importance of colocation between a real and a surrogate body for this to occur and thereby resolve a scientific controversy. This information is useful for exploiting immersive virtual reality in pain management.