Wave 2 · KHIDI RFP 11-1 [3-6] (p.61–64) · TRL 3 → 5 S-PUB ✓ PASS

Solo Reached: 85/100 · +CRO/Clinical: 94/100

3D Cardiac Dipole Patch-ECG Simulator & Real 48-Record MIT-BIH Ambulatory Arrhythmia Detector

48-Record MIT-BIH Patient-Disjoint Benchmark (109,966 Beats; DS2 AUROC 0.949 vs RR Baseline 0.870) + 3D Cardiac Dipole Patch Simulator

1. Frozen Inputs & SHA-256

Pinned

PhysioNet MIT-BIH Arrhythmia Database v1.0.0 RECORDS

Records: 48

SHA-256: fcdca7ead9fc93f612f7e025029241cb6429ba2fcd05327b661bdc6b26fbcb57

MIT-BIH 48 Ambulatory Records (.hea + .atr) + 12 Raw Format-212 360 Hz Waveforms (.dat)

Records: 108

SHA-256: e9084162fc6fe42e4a5589a9a15005796db8e2a3b5b882a900f69f6e536dfccf

Canonical Run Digest (runSha256)

37d3d94f8d22ca30eb0523de52a29e06dc96aea8566b94748b4aa6e0423908a8

AST Solver-Free Isolation

solverFree = True (0 violations)

Deterministic Replay CLI

PYTHONPATH=. python3 -m ecg11_36.orchestrate && PYTHONPATH=. python3 -m ecg11_36.audit && PYTHONPATH=. python3 -m unittest tests/test_ecg11_36.py

Download Audit JSON (ecg11_36-audit.json) ↗

Adoption Score Breakdown (0–100)

L1 Split & Protocol Lock	19 / 20
L2 First-Principles Sim	19 / 20
L3 Open Human Cohort	22 / 25
L4 Solver-Free Audit	15 / 15
L5 Statutory Readiness	10 / 20
1-Auditor Reached	85 / 100
+CRO / Clinical IRB	94 / 100

2. Solver-Free Recomputed Benchmark Metrics

Target: ANSI/AAMI EC57 patient-disjoint DS2 AUROC >= 0.92 & beats RR baseline

MIT-BIH DS2 Patient-Disjoint Ventricular Ectopy Detection AUROC (n = 330)

0.9493 AUROC

95% CI [0.9228, 0.9733]

Baseline Comparison (Naive Fixed-RR (<600 ms) Threshold Baseline)

0.8702 AUROC

Wearable Patch ECG Risk SaMD

Cardiologist-Annotated Ambulatory Beats Decoded (48 Records)

109966 beats

DS2 Raw 360 Hz Waveform Beat-Level PVC Specificity

98.06 %

DS2 Raw 360 Hz Waveform Beat-Level PVC Sensitivity

78.18 %

DS2 Raw 360 Hz Waveform Overall Beat Classification Accuracy

95.86 %

DS2 Patient-Disjoint Probability Calibration Brier Score

0.0805 Brier

DS2 Expected Calibration Error (10-Bin)

0.0855 ECE

Interactive Domain Simulation & Cohort Evidence Canvas

PASS

detectedBeatsInWindow

ectopicBurdenPct

0.04

snrDb

19.40

ventricularBeatsInWindow

Mathematical / Biophysical Methodology & Upgrade vs APOSTOLUS v1

Combines a McSharry-style 3D coupled dipole limit-cycle ODE and SNR < 6.0 dB motion-artifact abstention gate with binary WFDB .atr/.dat decoding across all 48 MIT-BIH ambulatory ECG recordings under ANSI/AAMI EC57 DS1->DS2 patient-disjoint evaluation.

Advancement Over Legacy APOSTOLUS v1

Replaces APOSTOLUS prevent12 textbook risk equations with real 360 Hz binary ambulatory ECG waveform/annotation analysis (PhysioNet MIT-BIH) + motion-artifact electrode impedance abstention.

3. Solver-Free Audit Gates (G1–G4)

✓ PASS

G1 · MIT-BIH 108-File SHA-256, Format-212 Binary & AAMI DS1/DS2 Split Lock✓ PASS
Verified all 48 .hea/.atr records (109,966 beats) and 12 Format-212 360 Hz .dat waveforms on disk, with strict 22-vs-22 patient-disjoint DS1/DS2 partitioning.
obs: {'ds1Patients': 22, 'ds2Patients': 22, 'errors': [], 'format212HeaderMismatches': 0, 'record100Beats': 2273, 'totalBeatsDecoded48': 109966}
G2 · McSharry 3D Cardiac Dipole Limit-Cycle & SNR < 6 dB Abstention Law✓ PASS
Confirmed RK4 unit-circle radial stability (<1e-3), physiological R-peak amplitude, and exact signal-to-noise recomputation enforcing HOLD at 3.8 dB.
obs: {'cleanRecomputedSnrDb': 18.5, 'errors': [], 'maxLimitCycleRadialDrift': 0.0, 'motionHoldRecomputedSnrDb': 3.8}
G3 · Solver-Free DS2 Ventricular Ectopy AUROC & Raw 360 Hz Beat Audit✓ PASS
Independently recomputed DS2 330-epoch ventricular ectopy AUROC vs fixed-RR (<600 ms) baseline and raw 360 Hz DS2 beat-level specificity/sensitivity.
obs: {'ds2BeatSensitivity': 0.781818, 'ds2BeatSpecificity': 0.980627, 'ds2BeatSpecificityCi95': [0.977761, 0.98313], 'ds2EpochAuroc': 0.949296, 'ds2EpochCi95': [0.922778, 0.973296], 'errors': [], 'fixedRrBaselineAuroc': 0.870204}
G4 · Electrode Detachment HOLD, AAMI Paced Quarantine & MFDS Hardware Gate✓ PASS
Preserved automatic SNR < 6.0 dB abstention, quarantined 4 AAMI-excluded paced records (102, 104, 107, 217), and documented domestic patch impedance bench requirements.
obs: {'errors': [], 'interactiveHoldCases': 1, 'pacedQuarantinedCount': 4}

Preserved Boundary & Negative Findings

F-11-36-SVEB-COMPENSATORY-BOUNDARY · Aberrated Atrial Premature Beats & Bundle-Branch Block Mimic Wide-QRS PVCs on Single-Lead Patches! BOUNDARY
Incorporating SA-node non-reset compensatory pause geometry raises DS2 epoch AUROC from 0.8702 (naive RR<600 ms) to 0.9493, while beat-level sensitivity on single-lead MLII holdout is 78.18% at 98.06% specificity due to RBBB/fusion overlap.
F-11-36-MOTION-DETACHMENT-HOLD · Ambulatory Electrode Pop & Stride Motion (SNR < 6.0 dB) Require Mandatory Abstention! HOLD
When patch contact degrades to 3.80 dB, high-slew electrode pop transients mimic polymorphic VT; automatic HOLD abstention prevents false emergency dispatch and prompts lead re-adhesion.
F-11-36-STATUTORY-PACED-AND-BENCH · AAMI EC57 Paced-Rhythm Hardware Spike Detection & MFDS Patch Bench Pairing Required◈ EXT
Per ANSI/AAMI EC57, 4 paced records (102, 104, 107, 217) are quarantined from unpaced VEB scoring because sub-millisecond pacemaker spikes require analog front-end detection and hardware impedance pairing prior to MFDS filing.

MFDS / IEC 62304 Dossier & Statutory External Gates

McSharry-Style Coupled 3D Cardiac Dipole ODE Morphology Synthesis
Patient-Disjoint Split Across MIT-BIH Ambulatory Holter Records (DS1 vs DS2)
SNR < 6 dB Ambulatory Motion-Artifact Automatic Abstention Gate
Ventricular Ectopy (PVC/VT) vs Supraventricular Sensitivity-Specificity Audit

Remaining External Statutory Gates (S-EXT)

ANSI/AAMI EC57 Arrhythmia Algorithm Beat/Rhythm Reporting & MFDS Wearable ECG SaMD Review
Korean Patch-Hardware Impedance Front-End Bench Pairing