Go To AIH-HSP Data Types

1 Definition 2 Functional Requirements 3 Syntax 4 Semantics 5    Conformance Testing 6     Performance Assessment

1 Definition

AIH Data Processing Types primarily serve to express inference of an individual’s health state. An AIH Data Processing Type identifies the kind of health state inferred from health data as a result of processing. The taxonomy applies to health data including behavioural, clinical, biological (omics), imaging, neurophysiological, and physiological data.

2 Functional Requirements

The Health Processing Type Taxonomy shall:

  • Provide unique, stable identifiers for health‑semantic processing types.
  • Describe processing in terms of health state inference, not computation methods.
  • Support both:
    • data‑type‑independent health inferences, and
    • data‑type‑specific health inferences where medically meaningful.
  • Enable TFA components to:
    • declare which health inferences they perform,
    • declare which health inferences they support,
    • document which health inferences have been applied to data.
  • Support extension with new processing types without breaking existing implementations.
  • Provide identifiers that are:
    • human‑readable,
    • machine‑processable,
    • version‑stable.
  • Provide a single authoritative namespace for health processing type identifiers within TFA.

3 Syntax

https://schemas.mpai.community/AIH1/V1.0/data/AIHDataProcessingType.json

4 Semantics

Label Description
HealthStateDetection Detection of the presence of a health‑relevant state or condition.
HealthStateClassification Classification of an individual into a health‑relevant state or category.
HealthStateQuantification Quantification of a health‑relevant measure or parameter.
HealthStateCharacterisation Derivation of characteristic properties of a health state.
HealthStateMonitoring Assessment of the evolution of a health state over time.
HealthStatePrediction Prediction of a future health state or outcome.
HealthPhenotypeDerivation Derivation of a health phenotype from health data.
BehaviouralStateDetection Detection of behaviour indicative of a health‑relevant condition.
BehaviouralStateClassification Classification of behavioural health states.
BehaviouralPatternCharacterisation Characterisation of behavioural patterns relevant to health.
BehaviouralTrendMonitoring Monitoring of behavioural changes over time.
BehaviouralRiskPrediction Prediction of health risk based on behavioural data.
ClinicalConditionDetection Detection of a clinical condition from clinical data.
ClinicalStateClassification Classification of an individual’s clinical state.
ClinicalOutcomeDerivation Derivation of clinical outcomes or endpoints.
ClinicalPhenotypeDerivation Derivation of a clinical phenotype.
ClinicalRiskAssessment Assessment of clinical risk or prognosis.
MolecularVariantDetection Detection of molecular or genomic variants.
MolecularProfileQuantification Quantification of molecular expression profiles.
MolecularSignatureCharacterisation Characterisation of molecular signatures.
BiomarkerStateDerivation Derivation of health state through biomarkers.
MolecularRiskPrediction Prediction of health risk based on molecular data.
AnatomicalStructureDetection Detection of anatomical structures or anomalies.
PathologicalStateClassification Classification of pathological imaging findings.
LesionExtentQuantification Quantification of lesion or structural extent.
ImagingPhenotypeCharacterisation Derivation of imaging‑based phenotypes.
DiseaseProgressionMonitoring Monitoring of disease evolution through imaging.
ImageDerivedRiskPrediction Prediction of health risk from imaging features.
NeuralEventDetection Detection of neurophysiological events.
BrainStateClassification Classification of brain or cognitive states.
NeuroActivityQuantification Quantification of neural activity measures.
NeuralPatternCharacterisation Characterisation of neural patterns.
NeuroStateMonitoring Monitoring of neurofunctional state over time.
NeurologicalRiskPrediction Prediction of neurological health risk.
PhysiologicalEventDetection Detection of physiological events.
PhysiologicalStateClassification Classification of physiological health states.
PhysiologicalParameterQuantification Quantification of physiological parameters.
PhysiologicalProfileCharacterisation Characterisation of physiological profiles.
PhysiologicalTrendMonitoring Monitoring of physiological trends.
PhysiologicalRiskPrediction Prediction of health risk from physiological data.

5     Conformance Testing

A Data instance Conforms with AIH Data Processing (AIH-HPT) if:

  1.  Its JSON Object validates against its JSON Schema.
  2. Any included  JSON Object validates against its JSON Schema.
  3. All Data in the JSON Object:
    1. Have the specified Data Types.
    2. Conform with the Qualifiers signaled in their JSON Schemas.

6     Performance Assessment

Go To AIH-HSP Data Types