Skip to main content
Figure 2 | Journal of Biomedical Semantics

Figure 2

From: Semantic enrichment of longitudinal clinical study data using the CDISC standards and the semantic statistics vocabularies

Figure 2

Mapping the CDISC ODM model to the data cube and DDI vocabularies. Illustrates how the original CDISC ODM model (depicted by rectangles in light gray) is overlaid with the RDF Data Cube (depicted by ellipses in green) and the DDI-RDF vocabularies (depicted by rounded rectangles in blue). The Data section, depicted on the left of the model, comprises a hierarchical structure whereby each level is fully contained within the preceding level. As the left side is more about structuring the clinical data, the Data section of the CDISC ODM model is more closely related to qb. The Clinical Data node is mapped to qb:Dataset while qb:Slice is used to split the Subject, Study Event and Form data nodes across the ODM hierarchy into slices, and the Item Data node is mapped to qb:Observation. The ODM node refers to the entire data set and is mapped to disco:LogicalDataset. The right side comprises the metadata section, which contains one Study node, which further comprises one MetaData node. The MetaData node contains a number of StudyEventDef, FormDef, ItemGroupDef and ItemDef nodes, one corresponding to each of the Subject, Study Event, Form, Item Group and Item data nodes defined in the Data section. The Metadata section shows how the variable definitions managed through disco matches ODM’s ItemDef while the grouping of variables via disco:Universe is applied at the FormDef level. Finally, Item Data is logically mapped to disco:Variable.

Back to article page