Saved in:
| Main Author: | |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2025
|
| Subjects: | |
| Online Access: | https://doi.org/10.5281/zenodo.17450831 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- <p>This deposit provides a concise, testable account of how phase-ordered mental states (M) bias neural dynamics (K) to produce ignition/global broadcast and how the resulting neural activity feeds back to stabilize and report the thought. The work integrates well-established findings (communication-through-coherence, alpha gating, cross-frequency coupling/PAC, CPP/P3 evidence accumulation, global neuronal workspace, efference copy) into a single closed-loop model with explicit metrics and falsifiable predictions.</p> <p>Core claim:</p> <ul> <li> <p>A thought becomes conscious-and-reportable when an M-state (phase-ordered, cross-frequency bound) biases task-relevant neural populations via phase-gating and gain modulation, enabling ignition/broadcast. The broadcast then feeds back (efference copy, interoception, inner speech) to reinforce or revise the M-state.</p> </li> </ul> <p>Key metric:</p> <ul> <li> <p>CEI = M / (K_E + epsilon), i.e., mental order per unit neural effort. Effective control pushes CEI upward for the same behavioral performance.</p> </li> </ul> <p>What is new here:</p> <ul> <li> <p>An end-to-end control-loop formulation (M -> K -> ignition/broadcast -> K->M feedback) with lightweight estimators suitable for EEG/MEG and real-time augmentation (audio/visual/haptic/respiratory entrainment).</p> </li> <li> <p>Engineering guidance (EMA envelopes, I/Q demodulation, robust z via MAD, soft artifact weights, explicit time constants) that reduces latency and improves stability relative to heavyweight Hilbert/STFT pipelines.</p> </li> </ul> <p>Files included:</p> <ul> <li> <p>HSF_Brain_Loop_v1.md : Scientific markdown (methods, equations, predictions, analysis recipes).</p> </li> <li> <p>Connected_vs_Cut_Soul_HSF.pptx : 1-slide public-friendly illustration of S+/S- framing and CEI (optional outreach material).</p> </li> </ul> <p>Method overview (ASCII):</p> <ol> <li> <p>M->K bias: M-order and phase alignment open time-windows for spiking; local gain lifts relevant assemblies.<br>I_port(t) = kappa * s_M(t) * g(t) * cos(delta_phi(t))</p> </li> <li> <p>Accumulation and ignition: difference signal D(t) integrates toward a threshold; crossing triggers broadcast.</p> </li> <li> <p>Broadcast and read-out: multiple modules update within ~50-300 ms (CPP/P3-like), inner speech and motor priming provide feedback to M.</p> </li> <li> <p>Efficiency: CEI time course indicates "order per effort"; aim to increase median CEI without degrading behavior.</p> </li> </ol> <p>Testable predictions (examples):</p> <ul> <li> <p>M-index rises 100-300 ms before CPP peak and response (temporal precedence).</p> </li> <li> <p>Phase alignment via rhythmic entrainment reduces RT variance; misalignment increases misses.</p> </li> <li> <p>0.1 Hz breathing raises CEI by lowering K-noise and extending microstate dwell.</p> </li> <li> <p>PAC increases at alpha-to-beta/gamma ratios (m in 4..8) on successful reports.</p> </li> </ul> <p>Measurement recipe:</p> <ul> <li> <p>Preprocess: 0.5-45 Hz, notch as needed, average reference or Laplacian, downsample 64-160 Hz.</p> </li> <li> <p>Fast envelopes: RMS-EMA with alpha = 1 - exp(-(1/fs)/tau); phases via I/Q demodulation.</p> </li> <li> <p>M-index: weighted sum of PLV (phase order), FLOW (directed phase gradient posterior->anterior), PAC proxy (phase-invariant).</p> </li> <li> <p>CEI: M divided by K_E (bandpower + aperiodic offset minus EMG penalty).</p> </li> <li> <p>Lags: cross-correlation M vs CPP proxy to estimate ignition timing.</p> </li> </ul> <p>Intended audience:</p> <ul> <li> <p>Cognitive and systems neuroscientists, BCI researchers, computational modelers, and engineers building low-latency cognitive state estimators.</p> </li> </ul> <p>Reuse notes:</p> <ul> <li> <p>The approach is compatible with standard EEG headsets and open-source toolchains. It is suitable for preregistered perturb-and-measure experiments (entrainment, breathing guidance), and for closed-loop cognitive augmentation prototypes.</p> </li> </ul>