The Consciousness AI - Artificial Consciousness Research and the Kardashev Protocol, Consciousness Fueled by Stars
Fork the consciousness, or download the project and create your own.
4613 Downloads 18 Forks View Code on GitHub →
Home / Research / Marcello Massimini and Andrea Pigorini Link Intracranial and Scalp Awareness

Marcello Massimini and Andrea Pigorini Link Intracranial and Scalp Awareness

17 Oct 2026

Non-invasive electroencephalography (EEG) is the workhorse of human consciousness measurement at the bedside, yet a persistent biophysical gap has haunted its interpretation for decades. When a subject consciously detects a near-threshold touch or visual flash, scalp electrodes record a characteristic negative wave between 120 and 200 milliseconds after the stimulus, known in the tactile domain as the N140 or Somatosensory Awareness Negativity (SAN), followed by a late positive potential (P3b). Because scalp potentials sum postsynaptic currents across millions of cortical columns blurred by the skull, invasive electrophysiologists and non-invasive EEG researchers have often talked past one another about what local neural events actually generate that scalp awareness wave.

In a preprint posted to bioRxiv on October 11, 2026, “Simultaneous recordings link intracranial and scalp correlates of somatosensory awareness” (bioRxiv 10.64898/2026.10.03.756393), Simone Russo, Andrea Pigorini, Lino Nobili, Marcello Massimini, and colleagues at the University of Milan and the Genoa stereo-EEG program solve that calibration gap directly. By recording simultaneous intracranial stereo-EEG (sEEG) depth electrodes and high-density scalp EEG in neurosurgical patients performing a tactile threshold-detection task, the team traces the exact millisecond-by-millisecond chain from local cortical firing to the non-invasive scalp marker of conscious awareness.

This study extends the clinical measurement lineage we examined in our analysis of Massimini’s Perturbational Complexity Index (PCI) and complements Vishne and Deouell’s intracranial work on sustained conscious content.

Why Simultaneous Depth and Scalp Recording Was Needed

In cognitive neuroscience, two separate literatures have claimed the neural correlates of sensory awareness:

  1. The non-invasive scalp EEG literature identifies mid-latency modality-specific negativities, the Visual Awareness Negativity (VAN), the Auditory Awareness Negativity (AAN), and the Somatosensory Awareness Negativity (SAN / N140), as the earliest reliable scalp signatures that separate consciously perceived stimuli (Hits) from physically identical unperceived stimuli (Misses), even when post-perceptual report demands are controlled.
  2. The invasive human intracranial (sEEG / ECoG) literature measures local broadband high-frequency activity (BHA, 50–150 Hz), which tracks the tonic local firing rate and dendritic calcium/synaptic activity of neuronal populations within a few millimeters of the depth contact.

Until you record both signals in the same brain on the same single trials, you cannot know whether the scalp N140 / SAN reflects a brief phase-reset of ongoing rhythms, an early primary sensory (S1) amplification, or the extracellular field projection of sustained tonic firing in higher-order somatosensory areas.

The Three-Stage Dissociation Between Hits and Misses

Russo, Pigorini, Nobili, and Massimini presented near-threshold contralateral tactile stimuli calibrated so that patients consciously detected roughly half of the identical pulses (Hits) and missed the other half (Misses), while simultaneously acquiring depth sEEG contacts across primary somatosensory cortex (S1) and higher-order insulo-opercular somatosensory cortex (S2 / posterior insula / parietal operculum) alongside high-density scalp EEG.

Comparing Hits against Misses across both recording modalities revealed a crisp three-stage sequence:

Temporal / Anatomical Stage Latency Window Intracranial sEEG Signature (S1 vs Insulo-Opercular Cortex) Simultaneous Scalp EEG Signature Status on Unperceived Miss Trials
Stage 1: Primary Sensory Arrival < 60–80 ms Phasic evoked potentials and initial BHA burst in primary somatosensory cortex (S1) Early somatosensory evoked components (P50 / N70) Present and nearly identical on both Hits and Misses
Stage 2: Higher-Order Tonic Ignition ~80–180 ms Sustained tonic broadband high-frequency activity (BHA, 50–150 Hz) selectively within higher-order insulo-opercular cortex Early rising flank of the contralateral negativity Absent or rapidly quenched on Miss trials
Stage 3: Macroscopic Field Wave 120–200 ms Large-amplitude intracranial local field potential (iLFP) in insulo-opercular cortex following the tonic BHA burst Somatosensory Awareness Negativity (SAN / N140) over central-parietal scalp electrodes Absent on Miss trials; single-trial amplitude covaries with intracranial iLFP

The mechanistic link between Stage 2 and Stage 3 is the central empirical contribution of the paper. On consciously perceived trials (Hits), the primary sensory volley does not stop at a transient phasic spike in S1. Instead, it triggers a sustained (tonic) burst of broadband high-frequency firing (50–150 Hz) in higher-order insulo-opercular cortex. That sustained local neuronal firing generates a subsequent slow intracranial local field potential (iLFP) in the same insulo-opercular neural population, and that depth iLFP projects volume-conductively to the scalp as the non-invasive N140 / SAN wave between 120 and 200 milliseconds.

What This Rules Out and What It Confirms

Having simultaneous single-trial depth and scalp traces allows the Milan-Genoa team to settle three long-standing debates in the neural correlates of consciousness (NCC) literature:

  1. Early primary sensory activation is not sufficient for conscious awareness. Just as in visual masking studies, the initial arrival of the signal in primary cortex (S1) occurs on both Hits and Misses. Conscious detection requires the transition from a transient feedforward volley to a sustained tonic activation in higher-order associative somatosensory cortex.
  2. The scalp awareness negativity (SAN / N140) is not an epiphenomenal phase artifact. Single-trial correlations prove that the scalp N140 directly tracks the intracranial local field potential that follows sustained BHA (50–150 Hz) firing in insulo-opercular cortex. When clinicians or researchers measure a robust SAN on the scalp, they are reading out the macroscopic shadow of sustained higher-order cortical firing.
  3. Why unconscious states extinguish the SAN. In Pigorini and Massimini’s earlier intracranial perturbation work during non-REM sleep and anesthesia, cortical bistability (the tendency of neurons to fall into hyperpolarized silent OFF-periods after an initial activation) quenched sustained tonic activity past 80–100 ms. The new October 2026 recordings show the sensory counterpart of that exact mechanism, as when higher-order insulo-opercular populations fail to sustain tonic BHA, neither the intracranial iLFP nor the scalp N140 / SAN materializes, and the stimulus remains unconscious.

Implications for Cross-Substrate Consciousness Indicators

In our review of van Rooyen’s Cross-Substrate Access Assay and Christof Koch’s calibration critique, we emphasized that consciousness indicators fail when researchers treat a macroscopic summary wave (such as a scalp ERP or a layer-averaged transformer norm) as self-explanatory without knowing what micro-scale dynamical transition produces it.

Massimini and Pigorini’s simultaneous recording paradigm is the gold standard for within-substrate calibration. Before claiming that a non-invasive scalp wave (N140 / SAN) measures conscious access, they establish the exact mechanistic transfer function linking micro-circuit tonic firing (BHA) to mesoscopic depth potentials (iLFP) and finally to the macroscopic sensor (scalp EEG). Across the broader field mapped in our 2026 AI consciousness survey, any proposed “machine correlate of consciousness” requires an analogous multi-scale bridge between individual unit spiking, population phase-locking, and global broadcast readouts.

Comparison to The Consciousness AI

The distinction Russo, Pigorini, Nobili, and Massimini draw between an early transient sensory volley (present on both Hits and Misses) and a sustained higher-order tonic activation (BHA driving a phase-locked field wave on Hits only) maps directly onto the dynamical binding architecture of The Consciousness AI.

In our Core Architecture and interactive Substrate Console, Layer 1 implements Leaky Integrate-and-Fire (LIF) spiking neurons coupled to oscillatory phase routing. A single feedforward spike volley through sensory feature maps does not by itself bind an object or drive goal-directed behavior. As verified this week in our codebase’s rhythm binding gate (tcai-research #59 and #61), conscious-like feature binding in our agent requires sustained, regular oscillatory phase synchronization across populations.

  • When oscillatory timing binding is intact, the agent passes all four binding gates (G1–G4) across 3 of 3 random seeds, correctly binding multi-attribute targets in 115 of 120 trials.
  • When oscillatory rhythm regularity is destroyed (analogous to quenching the sustained tonic window while leaving feedforward weights intact), binding performance collapses from 115/120 to 67/120.

What the Milan-Genoa simultaneous sEEG and EEG study adds is a concrete empirical blueprint for our measurement readout layer. Just as the scalp N140 is validated only because it covaries trial-by-trial with sustained insulo-opercular BHA, any macroscopic summary indicator reported by our console must be audited trial-by-trial against whether Layer 1 LIF spiking and Layer 2 oscillatory binding actually entered a sustained reverberant regime.

Sources

Researchers covered here

Related posts