The Clinical Inference Engine (CIE) explores how deductive, abductive (Bayesian), and inductive inference, causal models, and knowledge representation can support patient-specific reasoning from population-derived evidence. Clinical inference is fundamentally a process of iterative belief revision.
CIE owns patient-specific explanatory and probabilistic reasoning. It consumes
appropriate population-level causal knowledge, evidence, uncertainty, provenance, and
model versions from Models4PT without taking ownership of Models4PT's canonical
scientific knowledge. Physiological simulation research may inform that knowledge
through the separate hummod-research, JSim, Physiome, and Physiolog workflow.
CIE research outputs are hypothesis-generating until supported by separate clinical validation, uncertainty analysis, governance, and any applicable regulatory pathway. The repository must not contain patient data, credentials, or copyrighted local reference collections intended only for private study.
CIE is the canonical coordination center for the broader clinical inquiry program because patient-specific reasoning is its ultimate goal. Each connected repository remains authoritative for the content and systems it owns. Begin cross-repository work with the Clinical Inquiry Ecosystem Workspace briefing, which records the shared direction, project boundaries, current handoff, and links to authoritative sources.
Open ~/Projects/physiolog-simulations.code-workspace for model-development and
simulation work spanning six roots: Physiolog, hummod-research, the read-only HumMod
distribution, JSim, Models4PT, and CIE. This is distinct from the broader clinical
inquiry workspace described in the canonical briefing. Each root remains an independent
repository or external dependency with its own licensing, validation, and deployment
boundary.
This project is based in part on:
© 2026 Sean M. Collins Preprint available for personal and scholarly use at https://philpapers.org/rec/COLFPK