Hardware & streaming
- Wearable headband (Muse 2 / Muse S, 4 channels: TP9, AF7, AF8, TP10)
- Lab Streaming Layer (LSL) / Bluetooth into a Python real-time engine
- Causal SOS IIR filtering and online 2-second buffer windowing
Methodology · reproducibility · next phase
Both Semester 7 datasets failed the causal test for design reasons — no music in one, a fixed order in the other. The prospective wearable study below is built to remove exactly those confounds.
Experimental blueprint
Why counterbalancing fixes the PhysioNet confound
Semester 7 · PhysioNet (fixed order)
Every participant, same sequence — music condition = time on task.
| S1 | S2 | S3 | S4 | |
|---|---|---|---|---|
| P1 | None | Relax | Stress | AI |
| P2 | None | Relax | Stress | AI |
| P3 | None | Relax | Stress | AI |
| P4 | None | Relax | Stress | AI |
Semester 8 · Latin square (illustrative 4 × 4)
Each condition appears once in every serial position, decoupling music from practice and fatigue.
| S1 | S2 | S3 | S4 | |
|---|---|---|---|---|
| P1 | Silence | Pref. | Classical | AI |
| P2 | Pref. | Classical | AI | Silence |
| P3 | Classical | AI | Silence | Pref. |
| P4 | AI | Silence | Pref. | Classical |
Pref. = subject-preferred relaxing; Classical = standard classical; AI = AI-generated audio. Plus 3–5 min silent washouts between conditions.
Retain from Semester 7
Must be validated on real hardware
Source: reports/CONSOLIDATED_RESEARCH_FINDINGS.md · Section G
Pre-stated hypotheses
Neurophysiological modulation
Does personalized or relaxing background music attenuate the rise in frontal and temporoparietal theta/beta ratios during high-load N-back, compared with silence, under counterbalancing?
Behavioural performance
Does background music measurably improve d′ or reduce RT variability during sustained tasks once practice effects are controlled?
Subjective fatigue
Does auditory intervention reduce NASA-TLX workload and KSS fatigue over prolonged blocks compared with silence?
Calibration efficacy
Can a 2-minute pre-task resting recording match the single-trial classification performance of active task-based calibration (≈ 12.4 min)?
Reproducibility
Global random seed 0 across models, splitters, bootstraps and permutations. Raw data are git-ignored and verified by checksum.
$ python -m pytest tests/# Full unit-test suite (DSP, validation, PhysioNet, Phase 4 modeling)
$ python src/03_verify_baseline.py data/processed/features.csv# Historical MindWave baseline — 13 regression checks
$ python -m src.physionet.experiment1.behavioural_analysis# Behavioural pipeline & statistics
$ python -m src.physionet.experiment1.generate_sync_figures# Phase 3A synchronisation figures (5)
$ python -m src.physionet.experiment1.generate_preprocessing_figures# Phase 3B preprocessing figures (10)
$ python -m src.physionet.experiment1.generate_phase3c_figures# Phase 3C statistics & ML figures (12)
$ python -m src.physionet.experiment1.modeling.model_comparison# Phase 4 matched model evaluation & statistics
$ python -m src.physionet.experiment1.modeling.generate_phase4_figures# Phase 4 figures (9)
Test-suite status
PROJECT_RESEARCH_LOG.md §18 and phase4_final_validation.md §6.test_dsp.py, test_validation.py, test_phase9.py) — re-run while building this site.data/; without them they fail at load time.SHA-256 provenance
merged_eeg_datas_50_subjects.csv · 19,029,871 bytesfeatures.csvSHA256SUMS.txt · 706,091,119 bytesSynthesis
Master cross-dataset synthesis
Decisions register, protocol, reproducibility
MindWave
Claims boundary
Phase 8 ML (57 configurations)
Raw logged metrics
Friedman / Wilcoxon / alpha ablation
Raw audit & screening sensitivity
Test suite & leakage verification
predict_from_raw inference module
Headset pilot integration guide
PhysioNet Experiment 1
Phase 3C exploratory findings
d′, RM-ANOVA, practice confound
Dual-track DSP, 229-column features
Generic vs personalized modeling
Independent audit & trial reconciliation
Source: PROJECT_RESEARCH_LOG.md · §15, §18