Coral is a cohort analysis tool to interactively create and refine patient cohorts,
while visualizing their provenance in the Cohort Evolution Graph. The resulting cohorts can then
be compared, characterized, and inspected down to the level of single entities.
Getting Started
The workflow of Coral consists of two steps: creating cohorts, and characterizing them. Operations from these two categories are carried out in an iterative workflow.
Cohort Creation
An initial cohort that contains all items of the selected dataset is created automatically. Creation operations allow users to create new sub-cohorts based on different attributes and attribute combinations. Cohorts are refined with the Filter operation, or divided into multiple cohorts with the Split operation.
Cohort Characterization
Characterization operations give insights into the cohorts. Similarities and differences between cohorts can be checked visually with the View operation, and statistically with the Compare operation. Additional operations give access to prevalence information and the data of individual items.
Coral is developed by
Coral