Research Series
Governed Medical Digital Twins / MS-AGIP
A governed abductive architecture for medical Patient Digital Twins, operationalised for multiple sclerosis care in the MS-AGIP platform.
Research questions
- Why do LLM-based clinical explanations become fluent yet abductively unsound?
- Can abductive soundness — mechanistic plausibility, temporal coherence, explicit handling of missing premises and counter-evidence — be enforced architecturally rather than through training data alone?
- How should ontology-guided reasoning, a clinician-facing twin and a patient-controlled twin be separated at runtime?
Constituent works
-
PreprintConceptual Framework
From Plausible Narrative to Sound Abduction: A Governed Abductive Architecture for Medical Digital Twins in Multiple Sclerosis Care
Large language models (LLMs) embedded in medical Patient Digital Twins (PDTs) exhibit a systemic vulnerability: they can generate fluent, narratively persuasive, yet abductively unsound clinical explanations. In this context, abductive soundness means that an explanation preserves mechanistic…
Evidence status & planned validation
One conceptual-framework preprint (not peer-reviewed). Validation is proposed, not yet performed.
Manuscripts in this series build on a shared architecture; the abstracts and records above state each work's independent status.