Integrated Information Theory (IIT) predicts that the prefrontal cortex (PFC) is not required for consciousness.
The Cogitate study results regarding Question 1 supported Integrated Information Theory (IIT) overall, as conscious content was robustly represented in the posterior parietal cortex (PPC) and sensory regions, and only partly in the prefrontal cortex (PFC), though the absence of a conscious versus unconscious contrast prevented the isolation of strict neural correlates of consciousness.
The researchers will manipulate prefrontal cortex activity on a trial-to-trial basis using electrical stimulation in non-human primates and optogenetic silencing in mice.
Global Neuronal Workspace Theory (GNWT) requires that decoder performance in the posterior parietal cortex (PPC) after prefrontal cortex (PFC) manipulation is lower in mice and different in non-human primates (NHPs) to be considered a successful test result.
Integrated Information Theory (IIT) predicts that above-chance decoding of conscious content will persist in the posterior parietal cortex (PPC) and/or sensory regions during the stimulus onset window, even if the prefrontal cortex (PFC) is manipulated.
To test the Global Neuronal Workspace Theory (GNWT) prediction that conscious processing is represented by an ignition signal in the prefrontal cortex (PFC) and posterior parietal cortex (PPC) just after stimulus onset, the protocol assesses whether average decoding performance generalization is above chance level in the stimulus onset window, defined as 250–500 ms after stimulus onset for non-human primates (NHPs) and 150–400 ms after stimulus onset for mice.
To test the location of conscious content, the researchers will assess average decoding performance in the prefrontal cortex (PFC), posterior parietal cortex (PPC), and content-dependent sensory regions during the stimulus onset window (250–500 ms post-stimulus for NHPs; 150–400 ms post-stimulus for mice) and the stimulus duration window.
To enable cross-species comparisons, the study protocol involves non-human primates (NHPs) and mice viewing or listening to supra-threshold visual or auditory stimuli for variable durations in a go-nogo task that controls for report confounds, while recording neuronal responses from visual, auditory, posterior parietal, and/or prefrontal cortical areas using Neuropixels electrodes.
If causal manipulation of the prefrontal cortex induces inappropriate behavioral responses, the researchers will extend the manipulation duration from 0.1 seconds before sensory stimulus onset to the end of the behavioral response window.
The researchers will manipulate prefrontal cortex activity on a subset of trials to determine if the prefrontal cortex is required for stimulus decodability in the posterior parietal cortex and sensory regions.
In the testing protocol, a Bayes Factor (BF) greater than 10 is interpreted as strong evidence for hypothesis H1, which posits that decoding performance in the posterior parietal cortex (PPC) and/or sensory regions is greater than chance following prefrontal cortex (PFC) manipulation. A Bayes Factor less than 1/10 is interpreted as evidence for hypothesis H0, which posits that decoding performance is equal to chance after prefrontal cortex manipulation.
The researchers will test connectivity predictions by analyzing content-specific spike-LFP phase synchrony within and between the prefrontal cortex, posterior parietal cortex, and sensory regions to identify long- or short-range synchrony.
To test the Integrated Information Theory (IIT) prediction that conscious content is maximally encoded in the posterior parietal cortex (PPC) and/or sensory regions rather than the prefrontal cortex (PFC), researchers compare average decoding performance between these regions within the stimulus duration window.
Neuropixels probes are used to record spiking activity and LFPs simultaneously from two PFC areas, two sensory areas, and one parietal area in NHPs and mice during behavioral tasks.
Integrated Information Theory (IIT) predicts stronger decoding of conscious content in the posterior parietal cortex (PPC) and/or sensory regions than in the prefrontal cortex (PFC) during the stimulus duration window.
Global Neuronal Workspace Theory (GNWT) predicts increased connectivity from pre-stimulus to stimulus onset between the prefrontal cortex (PFC) and posterior parietal cortex (PPC), and between the PFC/PPC and sensory regions.
Global Neuronal Workspace Theory (GNWT) predicts increased functional connectivity between the prefrontal cortex (PFC), posterior parietal cortex (PPC), and sensory regions during ignition post-stimulus onset and offset.
Global Neuronal Workspace Theory (GNWT) predicts that conscious content is encoded in the prefrontal cortex (PFC) and the posterior parietal cortex (PPC) during the stimulus onset time window.
Integrated Information Theory (IIT) requires that decoding performance in the posterior parietal cortex (PPC) and/or sensory regions remains greater than chance following prefrontal cortex (PFC) manipulation to be considered a successful test result.
In non-human primates, the researchers will replace Neuropixels probes in the prefrontal cortex (areas 45/46) with linear electrode arrays (e.g., 24-contacts, 200 µm spacing) for stimulation across cortical layers after completing simultaneous recordings from PFC and posterior cortical sites.
Global Neuronal Workspace Theory (GNWT) predicts that decoding performance in the posterior parietal cortex (PPC) will be affected by prefrontal cortex (PFC) manipulation, with performance expected to be lower for mice and either higher or lower for non-human primates (NHPs), based on the inhibitory effect of the PFC manipulation.
Global Neuronal Workspace Theory (GNWT) predicts that decoding generalizes in the prefrontal cortex (PFC) and posterior parietal cortex (PPC) during ignition and continues as long as the stimulus is consciously perceived.
To test whether the prefrontal cortex (PFC) is required for consciousness, researchers assess whether average decoding performance differs during PFC manipulation trials compared to no-manipulation trials in the posterior parietal cortex (PPC) and sensory regions within the stimulus onset time window.
The neural predictions tested in this study are not unique to Global Neuronal Workspace Theory (GNWT) or Integrated Information Theory (IIT), as higher-order theories implicate the prefrontal cortex, while re-entry theories emphasize the role of the posterior parietal cortex and sensory regions.
Global Neuronal Workspace Theory (GNWT) predicts that decoding from the posterior parietal cortex (PPC) should be modified by prefrontal cortex (PFC) manipulation due to interference with ignition, while decoding from sensory areas should be preserved in the early time-window before ignition.
The researchers assess the location-based predictions of consciousness theories by testing whether activity in the prefrontal cortex, posterior parietal cortex, and sensory regions can decode task-irrelevant or non-target stimuli.
For Integrated Information Theory (IIT) to pass the experimental test, evidence must show that decoder performance is higher in the posterior parietal cortex (PPC) and/or sensory regions than in the prefrontal cortex (PFC).
Integrated Information Theory (IIT) predicts that decoding from the posterior parietal cortex (PPC) and/or sensory regions is not obliterated by prefrontal cortex (PFC) manipulation.
Global Neuronal Workspace Theory (GNWT) predicts that conscious stimuli are decodable from the prefrontal cortex (PFC) and posterior parietal cortex (PPC).
To test the Global Neuronal Workspace Theory (GNWT) prediction that an ignition signal occurs after stimulus offset, the researchers will assess whether the spike rate in the Prefrontal Cortex (PFC) and Posterior Parietal Cortex (PPC) during the stimulus offset window (250–500 ms after stimulus offset for non-human primates, and 150–400 ms after stimulus offset for mice) is greater than the baseline spike rate (0–250 ms before stimulus onset) using t-tests and Bayesian t-tests.
To test the timing of neural activity, the researchers will measure stimulus decoding in the prefrontal cortex, posterior parietal cortex, and sensory regions using a moving temporal window across the duration of the stimulus presentation.
For the Global Neuronal Workspace Theory (GNWT) to pass the test, decoding generalization must be greater than chance in the stimulus onset window in both the Prefrontal Cortex (PFC) and the Posterior Parietal Cortex (PPC).
Global Neuronal Workspace Theory (GNWT) maintains only a weak theoretical commitment to the requirement that spike rates be greater than baseline in the stimulus offset window in the prefrontal cortex (PFC) and posterior parietal cortex (PPC), because animals may not be consciously aware of the stimulus toward the end of the presentation window.
To account for trials where animals may not be consciously aware of the stimulus towards the end of the presentation window, the researchers will compare the average variance in spike rate in the stimulus onset window to the stimulus offset window in both Prefrontal Cortex (PFC) and Posterior Parietal Cortex (PPC) cells using an F test (α = 0.05).
The researchers will manipulate prefrontal cortex (PFC) activity from 0–0.5 seconds after stimulus onset in non-human primates and mice for irrelevant stimuli, using a single stimulus duration of 0.5 seconds.
The researchers will perturb prefrontal cortex (PFC) activity in a subset of recording experiments as a causal test to differentiate between Global Neuronal Workspace Theory (GNWT) and Integrated Information Theory (IIT).
A successful validation of Global Neuronal Workspace Theory (GNWT) in this study requires evidence that decoder performance is above chance in both the prefrontal cortex (PFC) and the posterior parietal cortex (PPC).
The Cogitate study results regarding Question 3 found no evidence for sustained short-range synchrony in the posterior parietal cortex (PPC) and sensory regions, but found partial evidence for long-range dynamic functional connectivity between the prefrontal cortex (PFC) and sensory areas, which supports Global Neuronal Workspace Theory (GNWT).
The Global Neuronal Workspace Theory (GNWT) requires that pairwise phase consistency increases to be greater than baseline during the stimulus onset window in the prefrontal cortex (PFC) and posterior parietal cortex (PPC) pair, the PFC and sensory regions pair, and the PPC and sensory regions pair to pass its testing protocol.
In the experimental protocol, prefrontal cortex (PFC) activity is manipulated in a subset of trials: non-human primates undergo electrical stimulation using a linear electrode array, while mice undergo optogenetic silencing via activation of channelrhodopsin-expressing inhibitory neurons using an optical fiber or focused laser.
The researchers plan to test Question 1 causally by directly manipulating activity in the prefrontal cortex (PFC) and assessing whether activity and stimulus decodability is modulated in the posterior parietal cortex (PPC) and sensory regions (a prediction of Global Neuronal Workspace Theory) or persists without being obliterated in the PPC and/or sensory regions (a prediction of Integrated Information Theory).