Neural correlates of perceptual consciousness from within: A narrative review of human intracranial research

  1. François Stockart
  2. Alexis Robin
  3. Hal Blumenfeld
  4. Milan Brázdil
  5. Philippe Kahane
  6. Liad Mudrik
  7. Jasmine Thum
  8. Michael Pereira
  9. Nathan Faivre  Is a corresponding author
  1. Department of Neurology, Yale University, United States
  2. Univ. Grenoble Alpes, Univ. Savoie Mont Blanc, CNRS, LPNC, France
  3. Neurology Department, CHU Grenoble Alpes, INSERM, U1216, Grenoble Institut Neurosciences, France
  4. Interdepartmental Neuroscience Program, Yale University, United States
  5. Department of Neurosurgery, Yale University, United States
  6. Department of Neuroscience, Yale University, United States
  7. 1st Department of Neurology, St. Anne Univ. Hospital and Faculty of Medicine, Masaryk University, Czech Republic
  8. Behavioral and Social Neuroscience Research Group, CEITEC MU, Czech Republic
  9. School of Psychological Sciences, Tel Aviv University, Israel
  10. Sagol School of Neuroscience, Tel Aviv University, Israel
  11. Brain, Mind, and Consciousness Program, Canadian Institute for Advanced Research (CIFAR), Canada
  12. Department of Neurosurgery, The University of Alabama in Birmingham, United States
  13. Univ. Grenoble Alpes, Inserm, U1216, Grenoble Institut Neurosciences, France
3 figures and 4 tables

Figures

Experimental approaches to study the NCC.

In the contrastive approach (left), the visibility of stimuli is manipulated across trials such that they are either perceived (conscious trials) or not perceived (unconscious trials). Two of the most popular paradigms to manipulate visibility, near-threshold presentation of stimuli and backward masking, are depicted. In report tasks, participants perform button presses or saccadic eye movements to report whether they perceived the stimulus or not. In no-report tasks, attentional manipulations, stimulus intensity, or automatic oculomotor or pupillary responses (e.g., optokinetic nystagmus, pupil dilation) are used to infer if the stimulus was perceived or not. Neural signals that correlate with consciousness should differ across trials that are classified as conscious and unconscious. In the supraliminal approach (right), full contrast stimuli are presented and are assumed to always be consciously perceived. Trials are compared across conditions that differ in the characteristics of the stimuli, such as presentation duration (discussed below).

Different types of implants used for intracranial recordings in humans: electrocorticography (ECoG) strips or grids are typically placed under the dura mater (purple implant); Deep-brain stimulation (DBS) electrodes are typically used to record and stimulate subcortical regions (green implant); Microelectrode arrays are positioned on the cortex (yellow implant); stereo-encephalography (sEEG) depth electrodes are inserted in the brain across cortical and subcortical structures (blue implant).

Hybrid models can record single-unit activity with either tetrodes (red) or microwires protruding from the electrode.

Summary of key results of intracranial studies of perceptual consciousness.

(A) Broadband activity in response to seen vs. unseen face stimuli in the prefrontal cortex follows a bimodal distribution, leading the authors to speculate that early responses may be associated with perceptual consciousness. (B) In the ventral visual cortex including the face fusiform area, High Gamma band Activity (HGA) increases track face stimulus detection in immediate and delayed report experiments, and stimulus intensity in a no-report experiment. (C) A thalamic awareness potential is observed in response to seen, but not unseen, face stimuli. Panel reproduced from Figure 2 from Kronemer et al., 2022. (D) Medial and intralaminar thalamic nuclei respond more to perceived face stimuli than ventral nuclei (left panel) and drive activity in the prefrontal cortex as shown by phase-amplitude coupling (PAC; right panel). (E) Multivariate decoding in the occipital and ventral-temporal cortices, but not the parietal and prefrontal cortices, tracks stimulus duration. Panel reproduced from Figure 4 from Vishne et al., 2023. (F) Stimulus category can be decoded for the entire duration of the stimulus in the posterior cortex but not in the prefrontal cortex, while stimulus orientation decoding does not track the duration of the stimulus in either region of interest. Panel reproduced from Figure 3 from Ferrante et al., 2025.

© 2024, Elsevier. Figure 3A was reprinted with permission from Figure 3 from Fang et al., 2024b which was published under a CC BY-NC 4.0.

© 2024, Springer Nature. Figure 3B was reprinted with permission from Figure 7 from Stockart et al., 2025a which was published under a CC BY-NC-ND 4.0. Further reproductions must adhere to the terms of this license.

© 2025, Science. Figure 3D is reprinted from Figures 2 and 5 from Fang et al., 2025, with permission from Science. It is not covered by the CC-BY 4.0 licence and further reproduction of this panel would need permission from the copyright holder

Tables

Table 1
Summary of electrocorticography (ECoG) and stereo-encephalography (sEEG) studies using the contrastive approach to study visual NCCs in humans.
StudyImplant typeCoverageSample sizeStimuliSuppression methodTask typeMain result
Brázdil et al., 2001sEEGTemporal and frontal13LettersNear-threshold (oddball task)ReportThe subliminal P3 peaks earlier and is shorter in duration than the supraliminal P3.
Naccache et al., 2005sEEGAmygdala3WordsBackward maskingReportMasked word emotion affected broadband activity ~870 ms after stimulus onset.
Gaillard et al., 2006asEEGOccipitotemporal7WordsBackward maskingReportRepetition of previously masked words exerts a long-lasting effect on neural signals.
Gaillard et al., 2006bsEEGOccipitotemporal1WordsBackward maskingReportCortical event-related response to masked words was confined to ventral pathways and not as sustained as for masked words.
Gaillard et al., 2009sEEGAll lobes (mostly posterior)10WordsBackward maskingReportPerceptual consciousness is associated with sustained activity in the prefrontal cortex and increased long-distance synchrony.
Fisch et al., 2009ECoGAll lobes (mostly posterior)11CategoriesBackward maskingReportStimuli reported as seen induce a non-linear increase of activity in higher-order visual cortex.
Aru et al., 2012aECoGOccipitotemporal6PicturesNear-thresholdReportA manipulation of visibility by prior exposure indicates that the higher-order visual cortex is not an NCC.
Vidal et al., 2014asEEGOccipitotemporal (ventral)3WordsBackward maskingReportRepetition suppression in the ventral visual cortex occurs for perceived and non-perceived stimuli.
Vidal et al., 2014asEEGAll lobes9Shape (ring)Contrast adaptationNo-reportStimulus disappearance triggered decreases in low-frequency activity and increases in HGA.
Baroni et al., 2017ECoGOccipitotemporal5FacesNear-threshold + continuous flash suppressionReportDecoding of activity in ventral visual cortex and lateral visual cortex correlates with subjective visibility better than with stimulus strength.
Haun et al., 2017ECoGOccipitotemporal6FacesNear-threshold + continuous flash suppression + backward maskingReportConscious contents show correspondence to integrated information measures based on neural activity.
Herman et al., 2019ECoG + sEEGAll lobes9FacesNear-thresholdReportPerceived stimuli, but not non-perceived stimuli, are accompanied by large-scale network switching.
Kronemer et al., 2022sEEGThalamus7FacesNear-thresholdReportA thalamic event-related potential with onset ~250 ms is observed in response to perceived vs. non-perceived stimuli.
Shan et al., 2022sEEGAll lobes7Not reportedBreaking continuous flash suppressionReportActivity from every channel could be used to discriminate between brain activity before and after stimuli broke from suppression.
Liu et al., 2023sEEGOccipital, temporal, parietal, and frontal13GratingsNear-thresholdReportClustering analyses showed an interaction between exogenous attention and conscious report.
Li et al., 2024ECoGVentral temporal4FacesNear-thresholdReportFace-specific activity in the ventral visual cortex is stronger and more reliable for perceived than non-perceived stimuli.
Fang et al., 2024asEEGMostly prefrontal6GratingsNear-thresholdReportStimuli reported as perceived induce prefrontal activations independently from motor preparation.
Fang et al., 2024bsEEGMostly prefrontal9GratingsNear-thresholdReportThe onset of broadband activity in response to perceived vs. non-perceived stimuli in the lateral prefrontal cortex follows a bimodal distribution.
Fang et al., 2025sEEGThalamus and prefrontal5GratingsNear-thresholdReportThalamic responses to perceived vs. non-perceived stimuli are earliest and strongest in medial and intralaminar nuclei, and drive lateral prefrontal cortex activity.
Stockart et al., 2025asEEGAll lobes29FacesNear-thresholdReport + no-reportA neural code in the ventral visual cortex reflects (1) stimulus detection and (2) stimulus intensity in a no-report experiment.
Table 2
Summary of electrocorticography (ECoG) and stereo-encephalography (sEEG) studies using the contrastive approach to study non-visual NCC in humans.
StudyImplant typeCoverageSample sizeStimuliSuppression methodTask typeMain result
Dykstra et al., 2016ECoGInferior temporal and inferior frontal1Tones (auditory)Informational maskingReportHeschl’s gyrus and the inferior frontal cortex increase their activity for heard vs. unheard sounds.
Christison-Lagay et al., 2025ECoG + sEEGAll lobes31Categories (auditory)Near-thresholdReportThe cortical and subcortical networks involved in auditory consciousness are similar to those in visual perception.
Albertini et al., 2025sEEGAll lobes30Median nerve stimulation (tactile)Near-thresholdReport + no-reportTonic HGA responses in the parietal operculum (1) are all-or-none, (2) differentiate between stimuli reported as perceived and not perceived, and (3) are observed in a no-report task.
Table 3
Summary of studies using the contrastive approach to study single-neuron NCC in humans.
StudyImplant typeCoverageSample sizeStimuliSuppression methodTask typeMain result
Kreiman et al., 2002MicrowiresMedial temporal14CategoriesFlash suppressionReportThe activity of neurons in medial temporal lobe follows reported percepts.
Quiroga et al., 2008MicrowiresMedial temporal5CategoriesBackward maskingReportThe activity of neurons in medial temporal lobe follows reported percepts.
Reber et al., 2017MicrowiresMedial temporal21Pre-screened stimuliAttentional blinkReportThe strength and timing of neurons’ responses to seen vs. unseen stimuli follows a posterior to anterior gradient in medial temporal lobe.
Gelbard-Sagiv et al., 2018MicrowiresMedial temporal and frontal9Pre-screened stimuliBinocular rivalryReportChanges in firing rates of medial temporal lobe neurons reflect the preferred stimulus start >1 s before a perceptual transition.
Pereira et al., 2021ECoG + arrayPosterior parietal1Vibrations (tactile)Near-thresholdReport + no-reportNeurons in the posterior parietal cortex reflect evidence accumulation, confidence, and stimulus intensity in the absence of reports.
Pereira et al., 2025MicroelectrodesSubthalamic nucleus + thalamus32Vibrations (tactile)Near-thresholdReportThalamic and subthalamic nuclei differentiate between perceived vs. non-perceived vibrotactile stimuli.
Vanhoyland et al., 2025ArrayLateral occipital cortex4CategoriesBackward masking + flash suppression + binocular rivalryReportClassification of stimulus category based on neural population responses only occurs for stimuli reported as seen.
Table 4
Summary of intracranial studies using the supraliminal approach to study the NCC in humans.
StudyImplant typeCoverageSample sizeStimuliMain result
Noy et al., 2015ECoGAll lobes43CategoriesContent-specific, high-magnitude visual cortex signals are followed by content-invariant frontoparietal signals.
Gerber et al., 2017ECoGOccipitotemporal10Faces and housesThe duration for which a stimulus is presented is better reflected by activity in posterior than anterior visual cortex.
Li et al., 2019ECoG + sEEGAll lobes11LettersPresentation of conscious stimuli leads to both sustained and transient dynamics in cortical network activity.
Kwon et al., 2021ECoG + sEEGAll lobes158WordsVisual, medial temporal, and frontal cortices form a signal detection network involved at the onset of perceptual consciousness.
Khalaf et al., 2023ECoG + sEEGAll lobes11LettersA large range of regions from all cortical lobes are involved in an early detection network.
Broday-Dvir et al., 2023ECoG + sEEGAll lobes13Faces and placesSustained representation of conscious stimuli depends on similarity distances between activation patterns in VVC.
Vishne et al., 2023ECoGAll lobes10Faces and housesThe population response in occipitotemporal regions encodes stimuli’s temporal dynamics.
Ferrante et al., 2025ECoG + sEEGAll lobes34CategoriesStimulus category can be decoded for the entire stimulus duration in the posterior region of interest, and after stimulus onset in the prefrontal region of interest.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. François Stockart
  2. Alexis Robin
  3. Hal Blumenfeld
  4. Milan Brázdil
  5. Philippe Kahane
  6. Liad Mudrik
  7. Jasmine Thum
  8. Michael Pereira
  9. Nathan Faivre
(2026)
Neural correlates of perceptual consciousness from within: A narrative review of human intracranial research
eLife 15:RP109604.
https://doi.org/10.7554/eLife.109604.4