talks and presentations
Click on POSTER to get a high resolution PDF file.
2025
- TalkDecoding the neural representations of conscious and unconscious perceptionPablo Rodríguez-San Esteban, Ana B. Chica, and José A. González-LópezReunión Científica sobre Atención (RECA) XIV Apr 2025
Thanks to advancements in machine learning, particularly in the field of deep learning, the past few years have seen a significant growth in brain-computer interfaces (BCIs) research, revealing their increasing applicability in neuroscience. This progress has expanded the scope of BCIs beyond clinical applications, revealing their potential in exploring fundamental aspects of neuroscience, including motor and cognitive processes. We will present data of an EEG study employing a detection/discrimination paradigm with Gabor stimuli. Using machine learning (ML) techniques, we were able to decode target presence and awareness with average AUCs of 0.75 at around 150 ms after stimulus presentation. These results demonstrated that the presence or absence of the target stimulus, as well as participants’ subjective perception, could be decoded from EEG shortly after target onset. We found some weak evidence of unconscious processing, as target presence could be decoded for unseen targets at both early and late time periods during the trial. Building on these findings, we are currently developing a closed-loop system capable of predicting perceptual lapsus by analyzing the participant’s EEG signal in real time, with the goal of improving task performance using alerting signals. The pipeline includes real-time preprocessing of the EEG signal, training a classifier, and deploying the model to continuously analyze the signal. In this presentation we will share some of the preliminary results we achieved when applying the closed-loop system, and outline the next steps for further development.
- PosterLa voz de la mente: Decodificación del habla a partir del EEG usando aprendizaje profundoJosé A. González-López, Owais Mujtaba Khanday, Pablo Rodríguez-San Esteban, and 2 more authorsIII Congreso Investigación PTS Granada Feb 2025
2024
- PosterClosed-loop neurofeedback to improve conscious perceptionPablo Rodríguez-San Esteban, Maarten Ottenhoff, Christian Herff, and 2 more authorsXIV Congreso de la Sociedad Española de Psicología Experimental (SEPEX) Oct 2024
Brain-computer interfaces (BCIs) are a robust and powerful tool that enables the integration of observed neural activity with direct manipulations of both this activity and behavior. BCIs have been widely used for years in clinical research, particularly with patients suffering from motor deficits caused by strokes or neurodegenerative diseases. Moreover, BCIs also have promising applications in basic neuroscience, with recent literature demonstrating their effectiveness in motor control, mental imagery, and attention. In a previous study we found that relevant information related to conscious perception processes can be decoded from electroencephalography (EEG) data using machine learning (ML) algorithms. Specifically, our results showed that the presence or absence of a stimulus could be decoded from the EEG data after target onset. Similar results were also obtained for participants’ reported perception. Furthermore, we also observed significantly different patterns of activation for present targets depending on whether they were reported by participants as seen or as unseen. Based on this perceptual task, we are developing a closed-loop system to detect perceptual failures (i.e. participants reporting a stimulus as ‘unseen’ when it was presented) to alert participants with an attentional signal in order to improve their performance in the task. This neurofeedback pipeline includes real-time preprocessing of the recorded EEG signal, training of a ML classifier, and deploying the model to continuously analyze registered data. Our closed-loop system is able to perform predictions of the performance on the behavioral task to decide whether to send an alert tone to the participant. Here, we present some preliminary results of this neurofeedback system, and lay the groundwork for future steps.
2023
- PosterDecoding visible and invisible perceptual stimuli from EEG data using Machine LearningPablo Rodríguez-San Esteban, José A. González-López, and Ana B. ChicaXIII Congreso de la Sociedad Española de Psicofisiología y Neurociencia Cognitiva y Afectiva (SEPNECA) Jul 2023
In recent years, the field of cognitive neuroscience has witnessed a remarkable surge in the utilization of machine learning (ML) techniques, particularly in the analysis of neuroimaging data. This burgeoning interest in ML models has significantly contributed to advancing our understanding of the inner workings of the human brain. Within this research line, we employed a task designed to study conscious perception. Participants had to discriminate the orientation of the Gabor lines (if the Gabor was not perceived they were asked to guess) and then indicate if they had consciously detected the appearance of the Gabor. Simultaneously, we recorded electroencephalography (EEG) data during task performance, utilizing ML algorithms to decode distinctive task-related parameters. Our preliminary findings demonstrate the feasibility of decoding various factors from EEG data, such as the presence or absence of the stimuli, participants’ reported perception of the stimuli, and the perceived orientation (clockwise or counter-clockwise). Moreover, we conducted a comparative analysis of the decoding performance when employing either raw voltage data or time-frequency features as input, revealing a substantial improvement in decoding accuracy when using time-frequency representations. Moving forward, our project will encompass further analyses aimed at generalization across different blocks and participants, as well as the adaptation of the analysis pipeline for real-time applications.
- PosterBoosting conscious perception through attention and neuromodulationPablo Rodríguez-San Esteban, Mar Martín-Signes, Cristina Narganes-Pineda, and 4 more authors26th meeting of the Association for the Scientific Study of Consciousness (ASSC) Jun 2023
Attention boosts perception of near-threshold stimuli thanks to the influence of fronto-parietal networks over sensory regions. Previous functional Magnetic Resonance studies in our group have demonstrated that phasic alerting and spatial attention improve perception through partially overlapping fronto-parietal regions, including the left Frontal Eye Field (FEF). In the present study, we exploit the power of Transcranial Magnetic Stimulation (TMS) to probe the causal role of the left FEF in the interactions between attention and conscious perception. Participants (N=50) detected near-threshold targets preceded by either auditory cues (to manipulate phasic alerting), or spatially predictive visual cues (to manipulate spatial attention). The behavioral task was performed after theta-burst stimulation to either the left FEF or the vertex (in separate sessions). Results demonstrated the expected attentional effects on reaction times and on the percentage of consciously detected targets (i.e., faster responses and better conscious detection for tone present vs. tone absent trials, and spatially valid vs. invalid trials). Moreover, response criterion was modulated by the stimulated region. TMS over the left FEF (as compared to vertex stimulation) altered response criterion for contralateral targets in the alerting task and for ipsilateral targets in the orienting task. Furthermore, previous studies in our lab have shown that white matter microstructural variability correlates with TMS effects. We hypothesize that TMS effects will correlate with the indexes of microstructural white matter properties of relevant tracts, such as the Superior Longitudinal Fasciculus and the Inferior Fronto-Occipital Fasciculus. Other data-driven approaches are envisioned (such as Tract-Based Spatial Statistics) to explore the role of additional tracts in the correlation between white matter integrity and TMS effects. Positive results would confirm the role of white matter variability in neuromodulation effects in the healthy brain.
- PosterThe role of white matter in neuromodulation effectsPablo Rodríguez-San Esteban, Joaquín J. Ramírez-Guerrero, Fátima Carmona, and 4 more authorsJornadas de Investigación 10º Aniversario CIMCYC: Pasado, Presente y Futuro May 2023
La atención es un mecanismo de selección y filtrado de información que permite potenciar el procesamiento perceptivo de estímulos difíciles de detectar, como los que se sitúan en el umbral de la consciencia. Estudios previos de nuestro grupo han encontrado que esta influencia de la atención sobre la percepción consciente está relacionada con la actividad de una red de regiones fronto-parietales que están conectadas anatómicamente mediante tractos de sustancia blanca, tales como el fascículo longitudinal superior, entre otros. En esta línea experimental estudiamos, mediante dos proyectos, cómo los efectos de la neuromodulación mediante estimulación magnética transcraneal (TMS) sobre regiones fronto-parietales como campo ocular frontal (FEF) o lóbulo parietal inferior (IPL) modulan el desempeño de los participantes en tareas atencionales y perceptivas y se relacionan con las propiedades microestructurales de la sustancia blanca. Ambos estudios emplearon paradigmas similares, donde 50 participantes sanos tenían que detectar estímulos Gabor en el umbral de la consciencia. Estos estímulos eran precedidos por claves auditivas (alerta fásica) o claves visuales (atención espacial). Para interferir en estos procesos se aplicaron diferentes protocolos de estimulación que se ajustaron a las características de cada tarea y sus objetivos. Los datos comportamentales de estos estudios se analizarán mediante 1) modelos lineales mixtos y 2) se correlacionarán con datos de sustancia blanca usando tractografía y el método conocido como Tract-Based Spatial Statistics. Estos resultados contribuirán a la comprensión del papel que juega la variabilidad microestructural de la materia blanca en los efectos de la neuromodulación en el cerebro sano.
- PosterMachine Learning approaches to investigate neurophysiological markers in language and perceptual tasksSneha Raman, Pablo Rodríguez-San Esteban, David Murcia, and 2 more authorsJornadas de Investigación 10º Aniversario CIMCYC: Pasado, Presente y Futuro May 2023
The use of machine learning (ML) techniques applied to cognitive neuroscience, especially for neuroimaging analyses, has seen exponential growth. A recent research line of our group focuses on the application of ML techniques to decode two different cognitive processes, perception, and speech, from neuroimaging data. In both lines, electroencephalography (EEG) was recorded from participants while performing either perceptual or language tasks. We applied ML algorithms to decode distinct parameters related to the task from the EEG data. For the perceptual tasks, participants had to detect near-threshold Gabor stimuli and report the orientation of said stimulus and whether they were able to see it or not. Preliminary results from the decoding analyses show that it is possible to discriminate these task responses from the recorded EEG data, with the peak decoding accuracy at around 500 milliseconds after stimulus presentation and in frequency bands such as alpha and beta. For the speech decoding tasks, we collected intracranial EEG data from epilepsy patients while they performed word recognition and picture naming tasks, where the participant produced spoken as well as imagined speech. The data went through a meticulous preprocessing procedure, which is now automatised. Preliminary evaluations suggest differences in high-gamma EEG features corresponding to spoken and imagined words, leading us towards gaining an understanding of brain regions involved in language production. Linear regression and deep learning algorithms were employed to decode speech from iEEG and to classify picture categories and the presence of vowels in the spoken words.
2022
- PosterSusceptibility to neuromodulation in the healthy brain: the role of white matter variabilityMar Martín-Signes, Pablo Rodríguez-San Esteban, Joaquín J. Ramírez-Guerrero, and 5 more authorsTranscranial Brain Stimulation in Cognitive Neuroscience Workshop Dec 2022
Transcranial Magnetic Stimulation (TMS)-induced neuromodulation is related to structural properties of white matter tracts. In patients with brain lesions, white matter properties have been linked with the recovery potential of TMS treatments. In healthy individuals, white matter microstructural variability also correlates with TMS effects. Tracts connecting the TMS-targeted region and other key regions are fundamental to explain TMS effects. Recently, some studies from our group have shown an association between TMS effects and indirect white matter tracts (not innervating the targeted region), probably due to compensatory processes supported by different anatomical networks. For example, phasic alerting produced a reliable behavioral boosting of conscious access, which was modulated by the stimulation of the supplementary motor area. This TMS effect correlated with the integrity of the right ventral branch of the Superior Longitudinal Fasciculus (SLF). Nevertheless, these (and similar) results have been observed with limited samples ( 20 participants). Currently, we are developing two projects in which we aim to replicate these findings with larger samples ( 50 participants). We are using a paradigm in which different attentional networks are manipulated (i.e., phasic alerting, and exogenous and endogenous orienting) while measuring its effects on the conscious perception of near-threshold targets. We hypothesize that TMS effects will correlate with the indexes of microstructural white matter properties of relevant tracts, such as the SLF and the Inferior Fronto-Occipital Fasciculus. Other data-driven approaches are envisioned (such as Tract-Based Spatial Statistics) to explore the role of other tracts in the correlation between white matter integrity and TMS effects.
- TalkLa dinámica cerebral en la integración de característicasMaría I. Cobos, Pablo Rodríguez-San Esteban, and Ana B. ChicaReunición Científica sobre Atención (RECA) XIII Oct 2022
Nuestro ambiente perceptual es complejo y tenemos la sensación de percibir una gran cantidad de información en cada momento. Sin embargo, nuestra percepción es limitada y a menudo sujeta a errores. Algunos de estos errores perceptuales consisten en integrar de manera incorrecta los rasgos de diferentes objetos, ocasionando lo que conocemos como ilusiones perceptuales. Este trabajo se compone de dos estudios que exploran la dinámica espacial (fMRI) y temporal (EEG) cerebral de estos fallos en la integración de características. Para ello, se usó una tarea dual en la que en algunos ensayos se producían errores de integración. En cada participante se manipuló el tamaño del estímulo objetivo con el fin de conseguir en torno a un 70% de respuestas correctas y un 30% de respuestas incorrectas. El estudio de resonancia mostró la implicación de regiones parieto-occipitales en el proceso de integración, con una mayor conectividad funcional para las respuestas correctas vs. incorrectas. El estudio de electroencefalografía mostró modulaciones en las diferentes frecuencias (theta, alpha, beta y gamma), tanto en momentos previos como posteriores a la presentación del estímulo. Ambos estudios revelan la participación de diferentes procesos cognitivos y cerebrales en las distintas fases de la integración de características.
- PosterLa integración de características en la dimensión espacial y temporalPablo Rodríguez-San Esteban, María I. Cobos, and Ana B. ChicaV Jornadas de Investigación CIMCYC: Cambio social desde la investigación basada en la evidencia May 2022
Las personas recibimos constantemente una gran cantidad de información, tanto del entorno como de nuestro propio organismo, y aunque tenemos la impresión de ser capaces de percibir la mayor parte de esta información, lo cierto es que sólo somos conscientes de una porción pequeña de ella. Esto es debido a que el proceso perceptivo es complejo y la capacidad de nuestros sistemas sensoriales es limitada, por lo que suelen cometerse fallos tales como las ilusiones perceptuales (errores en la integración de características de los objetos). Aquí se presentan dos experimentos que exploran las bases neurales asociadas a estos procesos, utilizando la electroencefalografía y la resonancia magnética funcional mediante una tarea dual. En esta tarea, el tamaño de los estímulos era manipulado con el fin de conseguir en torno a un 70% de respuestas correctas (correcta integración de rasgos) y un 30% de ilusiones (incorrecta integración de rasgos). En resonancia magnética observamos que hay un conjunto de regiones parieto-occipitales involucradas en la integración de características, mostrando mayor coactivación entre áreas parietales y occipitales para las respuestas correctas en comparación con los errores. En electroencefalografía se observan modulaciones en las frecuencias theta, alpha, beta y gama en momentos previos y posteriores a la presentación del estímulo. Estas modulaciones revelan la implicación de diferentes procesos cognitivos en la integración de características.
- PosterAtención y consciencia: sustancia blanca, variabilidad individual y neuromodulaciónMar Martín-Signes, Joaquín J. Ramírez-Guerrero, Pablo Rodríguez-San Esteban, and 1 more authorV Jornadas de Investigación CIMCYC: Cambio social desde la investigación basada en la evidencia May 2022
La atención es un mecanismo fundamental de selección y filtrado de la información que procesamos y que accede a nuestra consciencia. Mientras que la consciencia perceptual se asocia a un procesamiento neural distribuido en regiones caracterizadas por una amplia conectividad con módulos de procesamiento más sensoriales, la atención se relaciona con la actividad de una red fronto-parietal conectada anatómicamente mediante extensos tractos de materia blanca, entre los que destaca el fascículo longitudinal superior (SLF). En anteriores estudios de nuestro grupo se ha observado que las diferencias individuales en las propiedades microestructurales de la rama ventral del SLF se relacionan con los efectos neuromoduladores de la estimulación magnética transcraneal (TMS), aunque con muestras reducidas. Con el objetivo de aportar mayor consistencia a estos resultados y vislumbrar el valor predictivo que posee la sustancia blanca en los efectos de la neuromodulación, actualmente se están llevando a cabo dos estudios con muestras más amplias, que examinan la relación entre las características microestructurales del SLF y el efecto que la TMS produce en el comportamiento. Por último, un reciente proyecto del grupo está aplicando una nueva metodología de análisis que permite proyectar las activaciones de resonancia magnética funcional (fMRI) en la sustancia blanca para conocer su participación de forma directa. La utilización de datos de fMRI de paradigmas experimentales donde se manipulan los distintos sistemas atencionales y se mide la consciencia permitirá una novedosa aproximación al estudio de la sustancia blanca en estos procesos cognitivos.
- TalkParieto-occipital connectivity during correct and incorrect feature integrationPablo Rodríguez-San Esteban, Ana B. Chica, and Pedro M. Paz-AlonsoInternational APPE-SEPEX Meeting May 2022
Our sensory system constantly receives information from the environment and our own body. Despite our impression to the contrary, we remain largely unaware of this information and often cannot report it correctly. While perceptual processing does not require conscious effort on the part of the observer, it is often complex, giving rise to errors such as incorrect integration of features (illusory conjunctions). In the present study, we use functional magnetic resonance imaging to study the neural bases of feature integration in a dual task that produced around 30% illusions. A distributed set of regions demonstrated increased activity for correct compared to incorrect (illusory) feature integration, with increased functional coupling between occipital and parietal regions. By contrast, incorrect feature integration (illusions) was associated with increased occipital (V1-V2) responses at early stages, reduced functional connectivity between right occipital regions and the FEF at later stages, and an overall decrease in coactivation between occipital and parietal regions. These results underscore the role of parietal regions in feature integration and highlight the relevance of functional occipito-frontal interactions in perceptual processing.
2021
- TalkExplorando el papel de las áreas visuales y parietales en la integración de rasgosPablo Rodríguez-San Esteban, Ana B. Chica, and Pedro M. Paz-AlonsoXII Congreso de la Sociedad Española de Psicofisiología y Neurociencia Cognitiva y Afectiva (SEPNECA) Sep 2021
Nuestros sistemas sensoriales reciben constantemente una cantidad ingente de información del entorno y del propio organismo. Aunque tenemos la impresión de que somos capaces de percibir gran parte de esta información, lo que nos genera una experiencia subjetiva rica, lo cierto es que sólo somos conscientes, y podemos reportar de forma correcta, una pequeña porción de esta información. Esto se debe a que el proceso perceptivo, aunque no supone un esfuerzo consciente para el observador, es complejo y suelen cometerse fallos, tales como errores en la integración de rasgos. En el presente estudio utilizamos imágenes de resonancia magnética funcional para estudiar las bases neurales asociadas a este proceso de integración de rasgos, mediante una tarea dual que producía entorno a un 30% de errores en los participantes. Utilizando análisis de cursos temporales y de conectividad funcional, observamos que hay una serie de regiones parieto-occipitales involucradas en la correcta integración de rasgos. En primer lugar, las regiones visuales mostraron un incremento de la activación para los errores en comparación con los aciertos en etapas tempranas. Los análisis de conectividad, por su parte, revelaron mayor coactivación entre áreas parietales y occipitales para las respuestas correctas que para los errores. Estos resultados apuntan al importante papel que juegan en la integración de rasgos no sólo de las regiones parietales, algo ya reportado en la literatura previa, sino también de las áreas visuales, especialmente en los momentos más tempranos del proceso.
- Talk¿Lo ves? Efectos de la atención sobre la percepción conscienteMaría I. Cobos, Mar Martín-Signes, Pablo Rodríguez-San Esteban, and 1 more authorIV Jornadas de Investigación CIMCYC: Perspectiva de género en investigación psicológica May 2021
- TalkThe role of brain oscillations on illusory perceptionMaría I. Cobos, María Melcón, Pablo Rodríguez-San Esteban, and 2 more authorsJornadas Virtuales SEPEX 2021 Apr 2021
From all the information that reaches our senses, we can only consciously report a small part of it. However, we have the impression of perceiving much more information than we can report, which is known as phenomenological consciousness. Some authors argue that phenomenological consciousness is nothing more than a perceptual illusion according to which we believe to perceive more information than we can correctly report. In this study, we used high-density electroencephalography to explore the brain oscillations associated to correct or illusory perception using a dual task paradigm in which a percentage of perceptual illusions (erroneous conjunctions of features) are produced. Participants reported if a central digit was larger or smaller than 5, and then reported the color of a letter inside a peripherally presented string of characters. This string of characters included the target “L”, the distractor “O”, and two flankers. Behaviorally, all participants reported 30% illusions (reporting the color of the distractor), and these responses were comparable for the more or less demanding attentional conditions of the central task. Time-frequency analyses demonstrated that alpha (9-12Hz) power was reduced for hits as compared to illusions before the target was presented, demonstrating that attentional preparatory processes (related to sensory suppression) are associated to feature integration. Beta (17-23 Hz) power was also reduced for hits as compared to illusions, which might be related to more efficient attentional shifts for hits in the divided attention task used in this study. These results shed light on the brain mechanisms that dynamically interact to filter sensory information during visual perception.
- PosterParieto-occipital contributions to phenomenal consciousnessPablo Rodríguez-San Esteban, Ana B. Chica, and Pedro M. Paz-AlonsoJornadas Virtuales SEPEX 2021 Apr 2021
Our sensory system constantly receives information from the environment and our own body. Despite our impression to the contrary, we remain largely unaware of this information and often cannot report it correctly. Although perceptual processing does not require conscious effort on the part of the observer, it is often complex, giving rise to errors such as incorrect integration of features (illusory conjunctions). In the present study, we use functional magnetic resonance imaging to study the neural bases of feature integration in a dual task that produced ∼30% illusions. A distributed set of regions demonstrated increased activity for correct compared to incorrect (illusory) feature integration, with increased functional coupling between occipital and parietal regions. In contrast, incorrect feature integration (illusions) was associated with increased occipital (V1–V2) responses at early stages, reduced functional connectivity between right occipital regions and the frontal eye field at later stages, and an overall decrease in coactivation between occipital and parietal regions. These results underscore the role of parietal regions in feature integration and highlight the relevance of functional occipito-frontal interactions in perceptual processing.