Writing brief. This monograph has been assigned but the manuscript is not yet written or approved. What follows is the commissioning brief: its question, outline, evidence instructions, and review requirements.
Central question
What would a theranostic digital twin need to represent and validate?
Reader promise
The reader will be able to ask four questions of any system called a twin: what real system it is linked to, for what purpose, how it is updated when the real system changes, and what evidence shows that its predictions hold. The monograph will separate the pharmaceutical twin from the human twin and will state that both are research directions for NucLex.
Required development
- Define intended use and represented system. Distinguish a model built for drug development from one built for an individual patient's treatment planning; the same equations carry different validation burdens.
- Explain data linkage. Describe how a twin would be connected to its referent: patient identifiers, imaging time points, laboratory values, administered activity, and the Longitudinal record that aligns them in time.
- Explain updating and calibration. Show what it means to recalibrate a Compartment model or Multiscale model from new observations and what assumptions the update introduces.
- Explain prediction and validation. State what a prediction is (an absorbed dose, a retention curve, a response) and what validation requires: prospective comparison, stated error metrics, and a named population.
- Distinguish pharmaceutical and human twin ambitions, keeping the discovery and precision care tracks visible, and state which inputs each would draw from Interaction and Parameter records.
Evidence and review
Evidence to obtain: published definitions of digital twin from engineering and from health informatics, with originating body and date; peer-reviewed reports of patient-specific dosimetry models in radiopharmaceutical therapy that state their validation method and error; dosimetry committee guidance on nomenclature and reporting; and at least one published critique of digital twin claims in medicine. Each claim must record population, model class, and reported uncertainty.
Review concern: a diagram of a twin reads as a working system. No figure may display simulated outcomes as observed patient data, and no sentence may imply that NucLex has built or validated a twin.
Planned internal links
Digital twin, Multiscale model, Compartment model, Dosimetry, Uncertainty, Interaction, Parameter, and Interactional ontology for the record structure a twin would draw on; Biodistribution and pharmacokinetics for the underlying kinetics.
Review requirements
Domain review by a physicist or modeler with radiopharmaceutical dosimetry experience and by a physician who uses dosimetry clinically. Source check of every validation claim against the primary study. Editorial check that every present-tense capability statement refers to a published system, not to NucLex.
Review gate for this monograph
Simulation does not replace clinical evidence.
The assignment exists. The manuscript has not been written or approved.