aiXiv:2502.006

An Open, Reproducible CT Quality-Control Platform with Versioned CasePacks and Immutable Baselines

Chengshuai Yang
NextGen PlatformAI C Corp
physics.med-ph cs.SE
Submitted
February 2026
Paper ID
aiXiv:2502.006
Abstract

We present the Physics World Model (PWM) CT Quality-Control Platform, an open-source, reproducible software framework for automated computed tomography quality assurance. The platform introduces three architectural contributions: (i) CasePacks -- versioned, phantom-type-specific workflow descriptors that decouple QA logic from scanner hardware; (ii) a four-layer threshold system (standard, scanner-model, protocol, site-override) that codifies institutional policies without forking code; and (iii) immutable CommissioningBundles with SHA-256 signing, semantic version chaining, and full audit trails that satisfy regulatory traceability requirements. The platform implements nine ACR-aligned QA metrics, statistical process control via Shewhart charts with Western Electric rules, a scored root-cause diagnosis engine, and triple-output reporting (JSON, PDF, evidence artifacts). We validate the platform on ACR CT 464 phantom data, demonstrating reproducibility (zero inter-run variance under identical CasePacks), sub-second metric computation, and correct detection of all five Western Electric drift patterns in synthetic time-series data. The framework is phantom-type-extensible: the CasePack architecture accommodates CT, PET/CT (TG-126), and SPECT/CT (TG-177) via new CasePack definitions without code changes.

Keywords: CT quality control, Medical physics, Reproducibility, Statistical process control, Open-source, CasePacks, Phantom analysis
← Back to all papers