Skip to content

SPDX-FileCopyrightText: © Concepts 1996–2026 Miroslav Šotek.

SPDX-License-Identifier: AGPL-3.0-or-later

Commercial license available

© Code 2020–2026 Miroslav Šotek.

ORCID: 0009-0009-3560-0851

Contact: www.anulum.li | protoscience@anulum.li

SCPN Phase Orchestrator — Cardiac Rhythm Head-to-Head Benchmark

Cardiac Rhythm Head-to-Head Benchmark

This is the second deferred v1.0 domain benchmark page comparing the cardiac_rhythm domainpack against a cardiac-specific reference script.

Domain Mapping

  • Cardiac tissue is modelled as coupled phase oscillators with gap-junction coupling (approximation from Guyton, Ch. 10, and Strogatz's Kuramoto framing).
  • Layer structure maps pacemaking, atrial conduction, ventricular depolarisation, and repolarisation anatomy.
  • R_good targets healthy sinus/atrial synchrony (sa_node, atrial_conduction).
  • R_bad targets ventricular desynchronisation (ventricular_depolarization).

Benchmark Setup

  • File: domainpacks/cardiac_rhythm/run.py
  • Scenario:
  • sinus (steps 0..49)
  • PVCs (steps 50..99)
  • VT (steps 100..149)
  • drug (step 150..199)
  • pacing (steps 200..249)
  • Runtime: 250 steps, dt=0.002 (from domainpacks/cardiac_rhythm/binding_spec.yaml)
  • Metrics:
  • R_good: layers [sa_node, atrial_conduction] (indices [0, 1])
  • R_bad: layers [ventricular_depolarization] (index [2])

Head-to-Head Results

SPO benchmark runs (same scenario)

Mode Final R_good Final R_bad Final regime Notes
policy-disabled 0.6239 0.9999 nominal supervisor and policy actions disabled
policy-enabled 0.6105 0.9994 nominal default policy + supervisor

Baseline references

  • examples/cardiac_rhythm.py is a domain-specific 5-oscillator reference without policy/supervisor logic:
  • Normal sinus rhythm: R=0.997, 0.992, 0.988, 0.986 at 0.5 s intervals
  • AV block: R=0.997, 0.992, 0.987, 0.982
  • External pacemaker: R=0.964, 0.986, 0.892, 0.973
  • This baseline differs in network size and state composition from cardiac_rhythm, but preserves the same clinical interpretation: entrainment quality from pacemaker and pathology under conduction stress.

Repro Steps

PYTHONPATH=src python examples/cardiac_rhythm.py

PYTHONPATH=src timeout 240s python domainpacks/cardiac_rhythm/run.py

To reproduce the policy-disabled run for this scenario:

PYTHONPATH=src python - <<'PY'
import runpy
from scpn_phase_orchestrator.supervisor.policy import SupervisorPolicy
from scpn_phase_orchestrator.supervisor import policy_rules

SupervisorPolicy.decide = lambda self, upde_state, boundary_state: []
policy_rules.PolicyEngine.evaluate = lambda self, regime_id, upde_state, good_layers, bad_layers: []
runpy.run_path("domainpacks/cardiac_rhythm/run.py", run_name="__main__")
PY

Interpretation

  • cardiac_rhythm remains in nominal regime in both settings over this scenario.
  • With the default supervisor, the run keeps R_good near policy-off execution while suppressing R_bad marginally.
  • Baseline and full-domain setups are complementary; together they cover a small closed-loop orchestrator against a domain-focused entrainment toy reference.