TwinRehab · In Silico Trial — IRHIS

Synthetic patient model cards

Every digital twin in this trial is a documented model, not an improvised profile. These cards state, for each seed model, exactly what it represents, which assumptions it rests on, how it evolves over time — and where each number comes from in the clinical literature.

VERSION v1.0 STATUS Public model card SCOPE Level 1 — technical/analytical performance SEED MODELS 2 of 90-twin cohort

Synthetic data, for software validation only. These models generate realistic data flows to test that the platform transmits and displays values faithfully. They are not clinical evidence and do not predict any individual's recovery: clinical performance is evidenced by the real-patient arm. Trajectories describe a typical course and forward values are indicative.

The seed models

Two twins, fully documented

These two models are the seeds from which the full cohort of 90 digital twins is generated, by varying joint, phase, sex, age, baseline severity, adherence and recovery archetype.

Common data schema

What each twin emits

Both models share one schema, mapped onto HL7 FHIR resources, so the synthetic flows speak the same language as the future operational platform.

How the numbers are produced

Generation logic & reproducibility

Evidence base

Literature anchors

Every value in these cards is traceable to one of the sources below. Superscript numbers in the rationale sections refer to this list.