FHIR utility for data science and IA

Hello community,

I have a question that is general, what the utility of FHIR for IA projects?
in other terms , if I have my data standardized FHIR what would make me
advantageous than someone who doesn’t have his data standardized FHIR concretely.

#IA #Data science
Thank you!

You’d have standardized data :slight_smile: If you’re looking to gather data from a variety of sources, it’s easier to run an AI algorithm if all of that data is structured in the same way. Otherwise, you need a distinct algorithm per data source or you need to go through a data normalization step before you can run your algorithm.

In terms of “Why FHIR vs. other syntaxes?”, FHIR data tends to be someone more consistent when you get it from multiple sources than something like HL7 v2, CDA, etc. That said, while FHIR can support pretty much any kind of data that’s even tangentially related to healthcare your algorithms might be interested in, that doesn’t mean all source systems you wish to acquire data will be able to express all of those data elements using FHIR.


That’s very clear, thank you for your answer.