The Smart CDSS
Clinical Decision Support System for Cancer


HL7 vMR
Arden Syntax
HL7 FHIR

Standardization > HL7 vMR
HL7 vMR

  • HL7 Virtual Medical Record (vMR) represents patient data in standard data model which is relevant to CDSS interventions.
  • Moreover, it also defines standard interfaces by mentioning standard input and output.
  • HL7 vMR defines Domain Analysis Model (DAM) which support following specifications.
  • Data model specification for CDSS.
  • Structured specifications for input and output of the CDSS engine.
  • Structured specifications for defining requirements of the input of CDSS intervention.



Standardization > HL7 Arden Syntax
HL7 Arden Syntax

  • The Smart CDSS knowledge base contains rules published by clinicians in HL7 Arden syntax.
  • Each rule constructs some recommendations, alerts, or guidelines, represented in MLMs.
  • These modules have well-defined structures to represent rules.
  • For example, the ¡°maintenance¡± section defines the title of the MLM, the author of the MLM, and the date and time of the creation of the MLM.
  • The ¡°library¡± section contains information that describes the purpose of the MLM.
  • The ¡°knowledge¡± section holds the data and logic of the MLM that will generate recommendation or alerts.



Standardization > HL7 FHIR
HL7 FHIR

  • To keep align the view derived from common reference model, FHIR based standard resources are bundled according to requirements of query in MLMs.
  • FHIR based resource view generation from MLM¡¯s query is performed in two steps:
  • vMR data model is mapped into available FHIR resources.
  • As FHIR is evolving, so for unmapped resources, vMR concept is converted into resource.
  • FHIR based profile is generated from query for each view, that helps in data integration expert to implement appropriate end point.
  • Providing FHIR based resource view with associated profile gives understandable interface requirements for data integration experts and allows knowledge expert to work transparently without knowing technicality of diverse data sources.