Cardiac Ventricular Late Potentials (SAECG) Skill

Zero-dependency sovereign personal agent skill for non-invasive evaluation of Signal-Averaged ECG (SAECG) Ventricular Late Potentials (VLPs), terminal microvolt fragmentation, and arrhythmogenic substrate screening.

flowchart TD
    RawECG["High-Resolution Multilead Signal-Averaged ECG (>= 250 beats)"] --> BiFilter["Bi-Directional 40 Hz High-Pass Butterworth Filter"]
    BiFilter --> VectorMag["Vector Magnitude QRS Complex: X^2 + Y^2 + Z^2"]
    VectorMag --> SimsonMetrics["Simson Criteria: Filtered QRS (fQRS), RMS40 (uV), LAS40 (ms)"]
    SimsonMetrics --> VLP_Check["Late Potentials Gating (>= 2 of 3 Criteria Met)"]
    VLP_Check --> SynchronyScore["Myocardial Conduction Synchrony & Arrhythmic Risk Score"]
    SynchronyScore --> Agent["OpenClaw / Hermes Personal Agent Recovery Guarding"]

Features

• Deterministic Math: Pure Python standard library implementation with zero external dependencies.
• Electrophysiology Standards: Implements validated Simson 40 Hz high-pass criteria for ventricular late potentials.
• Myocardial Electrical Stability: Highly sensitive detection of slow, fragmented conduction predisposing to re-entrant ventricular tachyarrhythmias.
• Model Context Protocol (MCP): Native JSON-RPC tool integration for LLMs and autonomous agents.

Quickstart

from client import CardiacVentricularLatePotentialsClient

client = CardiacVentricularLatePotentialsClient()
result = client.evaluate_saecg(
    filtered_qrs_duration_ms=98.5,
    rms_40_uv=38.2,
    las_40_ms=24.0
)
print("VLP Positive:", result["vlp_positive"])
print("Synchrony Score:", result["myocardial_synchrony_score"])
print("Classification:", result["classification"])

License

Apache-2.0