Physiological digital twins require cross-system observation contracts
Whole-body models explain how organ systems can be represented and coupled. A model that is updated repeatedly from patient measurements must also specify when those measurements support a particular conclusion. This Hypothesis and Theory article proposes an Observation Contract Framework: a declared set of evidence requirements for each output, checked whenever that output is requested. Each system uses 10 fields covering target states, direct measurements, proxies, unmeasured states, events, reference measurements, timing, quality failures, prohibited conclusions and measurement readiness. The framework returns permit, degrade or block for each output. Five testable hypotheses address unsupported conclusions, timing errors, missing reference measurements, cross-system dependencies and the distinction between signal reconstruction and state validation. The framework is specified for eight peripheral systems, with central nervous system dependencies explicitly left unobserved. A mathematical oxygen-delivery example explains how system requirements are combined. A small retrospective replay of public critical-care records then illustrates the same rules: a record-based oxygen-delivery proxy can be restricted or withheld as haemoglobin measurements become outdated, while a validated physiological oxygen-delivery claim remains blocked. This replay is an application example, not a clinical validation or a comparative test of the hypotheses. Existing sensors support many bounded measurements, but do not continuously identify regional perfusion, venous capacitance, ventilation–perfusion mismatch, endocrine pulsatility, tissue immune activity or in vivo coagulation potential. The framework complements context-of-use definition, observability and identifiability analysis, model interoperability, verification, validation, and uncertainty quantification. Its purpose is to state what evidence is needed for a conclusion and what must be withheld when that evidence is insufficient.
Authors
- Xingshuo Dong (ORCID: https://orcid.org/0009-0004-2601-968X)
Institutions
- Chinese PLA General Hospital (CN)
- Chinese People's Liberation Army (CN)
Publication Details
- Journal
- Frontiers in Physiology
- Published
- 2026-11-20
- DOI
- https://doi.org/10.3389/fphys.2026.1947699
- Primary Topic
- Healthcare Technology and Patient Monitoring
- Type
- article
- Field-Weighted Citation Impact
- 0.00