Open einargs opened 11 months ago
I took the data from the resources and loaded them into a dictionary. There was a lot of unnecessary info in the dictionaries so I cleared them out from dictFile.txt. I was working on publishing dev-summarize branch to the repo.
This will be the meat of the project. Converting the classes from the FHIR library into formatted data that the chatbot can easily digest and understand.
Symone Randle and Ermias are currently set up to work on this, though other people can join in and help.
What we need built is a function (that can call other functions) that takes a Bundle and gives us a string of formatted text describing important information about each entry in the bundle.
We also need you to ignore records that aren't important, like Claim or ExplanationOfBenefit.
You may need to do some work to e.g. resolve resource links. One possible way of doing this is to just store all bundle entries in a python dictionary indexed by their URL/identifier.
So for a patient record it might do:
We have synthetically generated patient records in the
data/fhir
directory. They use the FHIR R4B format.You can use locally hosted pydoc to look at the classes in the
fhir.resources
library. More information is in the project README on github: https://github.com/einargs/tnhimss-bill/Also useful is the FHIR specification: https://www.hl7.org/fhir/ I normally just google FHIR + whatever resource type I need to know about.
The repository has an example JSON file downloaded from synthea that is a bundle of FHIR records associated with a single patient. It can be very useful to ctrl-f through as you’re reading through FHIR documentation, since each node says what class it is: https://github.com/einargs/tnhimss-bill/blob/main/data/fhir/Aaron697_Brekke496_2fa15bc7-8866-461a-9000-f739e425860a.json
This is a zip file with a bunch of FHIR R4 synthetically generated health records from synthea. You can just download these, stick them in the data/fhir directory, and test with them. https://synthetichealth.github.io/synthea-sample-data/downloads/synthea_sample_data_fhir_r4_sep2019.zip