Article — Position paper · ○ Open access

The patient digital twin: anatomy of an epistemological misunderstanding

From the single model to the digital twin generator

Jérôme Vetillard · · Twingital Institute · 1 min read
🇫🇷 Lire en français ↓ Download PDF

The epistemological confusion

Four levels that the HealthTech industry routinely conflates: statistical model, digital shadow, digital thread, and digital twin. Approximately 90% of what is marketed as a “digital twin” is, in reality, a statistical model.

The reductionist bias

The illegitimate scale crossing — from molecular to cellular to tissue to organ to organism to clinical — is what most architectures commit. Each transition involves an epistemological leap that cannot be bridged by mere data integration.

The HDLSS chasm

Four consultations per year. Less than 1% data coverage. Missing Not At Random data mechanisms. This regime is structurally incompatible with industrial-grade twinning.

The generator thesis

The answer is not a better twin but a generator of digital twins. That is TweenMe — a platform that generates domain-specific twins from composable ontologies.

Read the document