Architecture

DT4H is a layered intelligence architecture for Digital Twin for Health systems.

The architecture separates signals, cohort intelligence, reference-human priors, Twin initialization, calibration, StateK computation, and SETPOINT execution.

Architecture diagram #

INPUTSSignals

Biomarkers · behavior · clinical · self-report

DT4HModeling Infrastructure

Cohorts · reference humans · Twin initialization · calibration

STATEKState Computation

State · confidence · trajectory · transition logic

SETPOINTExecution Runtime

Protocols · practices · outcomes · feedback

Layer responsibilities

DT4H is designed as an infrastructure layer, not an end-user wellness app. Each layer has a bounded responsibility so that modeling, state computation, and execution remain separable.

01

Signal Layer

Normalizes biomarkers, behavior, clinical context, and self-report.

02

Cohort Layer

Resolves probabilistic population context for inference and initialization.

03

Reference Layer

Provides distributional priors, expected ranges, and variance envelopes.

04

Twin Layer

Creates and maintains the individualized longitudinal model.

05

StateK Layer

Computes state, confidence, trajectory, and transition readiness.

06

Execution Layer

Feeds SETPOINT protocols, practices, outcomes, and recalibration.

System contract

InputSignals

Evidence enters the platform as normalized runtime signals.

ModelingDT4H

Cohorts, reference humans, initialization, and calibration.

ComputationStateK

Current state, confidence, transition logic, and readiness.

ExecutionSETPOINT

Protocols, practices, outcomes, and feedback.

Boundary principles

01

Modeling is not diagnosis

DT4H outputs are computational modeling signals, not clinical conclusions.

02

State is not action

StateK computes readiness and transition logic, while SETPOINT handles execution.

03

Execution feeds learning

Outcomes from protocols and practices return as evidence for recalibration.

LayerArchitecture Layer
StatusActive Draft
SystemDT4H / StateK / SETPOINT
BoundaryInfrastructure, not diagnosis
System lineageDT4HTwinStateKSETPOINTOutcomesRecalibration
Infrastructure boundaryDT4H models cohorts, Twins, calibration, and runtime state. It does not diagnose, prescribe, or replace licensed clinical judgment.
Document statusInfrastructure draft
Last updatedMay 2026
Applies toDT4H.ai / AvatarK.ai ecosystem