Skip to content
All posts

Lynx Analytics and Katana Graph demonstrate that Electronic Health Record data can be used to predict patient journeys

We recently exhibited at the BioTechX conference in Basel Switzerland with our partners, Katana
Graph. Using our combined technologies, we showed how to use graph representation learning to
predict patient outcomes following admission to an Intensive Care Unit (ICU).

This was done using data from patients’ Electronic Health Records (EHRs). Details of the data set, methodology, and results can be found here.

Using LynxKite and Katana’s Graph Intelligence Platform, we were able to reproduce in a few days
results from a research article that originally took 7 years to complete.

This demonstrated the power of LynxKite to quickly build graph models and get computationally intensive results in a short amount of time.