Stream: connectathon mgmt
Topic: Storage and Analytics track
Alexander Zautke (Apr 16 2018 at 13:21):
Hi everyone,
based on the initiative of @nicola (RIO/SS) there will be a Storage and Analytics track at the connectathon in Cologne.
We wanted to kick off the discussion about the exact scenarios of this track.
Anyone, who is interested, is more than welcome to discuss and share any thoughts.
For more information and the current planning, please see:
http://wiki.hl7.org/index.php?title=201805_FHIR_Storage_and_Analytics
nicola (RIO/SS) (Apr 16 2018 at 15:04):
:thumbs_up:
nicola (RIO/SS) (May 09 2018 at 16:27):
Let's work on orientation presentation - https://docs.google.com/presentation/d/1lSHzqZa4V9dU-oM3PDZGnZ2GR6MZK1agpyPSRVlly00/edit?usp=sharing. @Alexander Zautke @Tim Zallin
nicola (RIO/SS) (May 09 2018 at 16:34):
For connectathon we need dataset to play with storage & queries - we can use mitre https://syntheticmass.mitre.org/download.html. Any other ideas?
Alexander Zautke (May 09 2018 at 16:44):
In my opinion we should try to decide what the overall goal of this track is in order to reduce our current scope a bit. What do we want to achieve?
nicola (RIO/SS) (May 09 2018 at 16:49):
First of all get together - all of us, who implementing storage, fhir search & analytics on top of FHIR and have many ideas & concerns.
nicola (RIO/SS) (May 09 2018 at 16:50):
This is the fist-time track for this topic - so may be good result of it to define it's scope better for next connectathons :)
Alexander Zautke (May 09 2018 at 16:57):
Great :) Just wanted to know, if you had any particular outcome in mind.
David Hay (May 10 2018 at 08:18):
Just remember the wifi issues - if there's any way of organizing a local server with data, that may be optimal....
nicola (RIO/SS) (May 11 2018 at 07:36):
Do we have local servers?
nicola (RIO/SS) (May 12 2018 at 08:42):
Hey guys here is the repo - https://github.com/fhir-fuel/fhir-storage-and-analytics-track
nicola (RIO/SS) (May 12 2018 at 08:42):
Send me your github accounts
Toby Hu (May 12 2018 at 08:46):
Toby: https://github.com/toby-hu
nicola (RIO/SS) (May 12 2018 at 08:46):
https://github.com/fhir-fuel/fhir-storage-and-analytics-track/invitations
nicola (RIO/SS) (May 12 2018 at 08:47):
I'm gonna create a template readme so we will fill it with specific technology
nicola (RIO/SS) (May 12 2018 at 08:47):
This is sample subfolder, which i started for postgresql - https://github.com/fhir-fuel/fhir-storage-and-analytics-track/tree/master/postgresql
Christiaan Knaap (May 12 2018 at 08:48):
should we create a specialized stream for this? Or should we join the existing 'analytics on fhir' stream?
nicola (RIO/SS) (May 12 2018 at 08:48):
Let's join existing one
Alexander Zautke (May 12 2018 at 08:49):
Alex: https://github.com/alexzautke
nicola (RIO/SS) (May 12 2018 at 08:51):
Alexander you are already there
nicola (RIO/SS) (May 12 2018 at 08:57):
Here is basic template - https://github.com/fhir-fuel/fhir-storage-and-analytics-track/blob/master/TEMPLATE.md
Fill free to extend and fix it :)
Yatish Gupta (May 12 2018 at 09:15):
Send me your github accounts
nicola (RIO/SS) (May 12 2018 at 09:28):
Send me your github accounts
done
nicola (RIO/SS) (May 12 2018 at 09:42):
Let's use FHIR from SyntheticMass Data, Version 2 (24 May, 2017) - https://syntheticmass.mitre.org/downloads/2017_05_24/synthea_1m_fhir_3_0_May_24.tar.gz as a dataset
nicola (RIO/SS) (May 12 2018 at 11:29):
Tim Zallin (May 12 2018 at 11:54):
https://uhn-aidbox.health-samurai.io/Patient?__secret=verysecret&_count=1
Tim Zallin (May 12 2018 at 11:55):
some examples of sql queries available http://uhn-jupyter.health-samurai.io/notebooks/postgres_examples.ipynb
Tim Zallin (May 12 2018 at 11:56):
token is b95c0b839af782e11c13d55e1c73de2cb4c82158040948f1
Tim Zallin (May 12 2018 at 11:59):
attribute example https://uhn-aidbox.health-samurai.io/Attribute?__secret=verysecret&_count=1
nicola (RIO/SS) (May 12 2018 at 12:01):
Here is freshly generated synthea data public link - https://storage.googleapis.com/fhir-storage/synthea.tar.gz
Yatish Gupta (May 12 2018 at 12:13):
Thanks! Do we have a list of analysis we want to do?
Alexander Zautke (May 12 2018 at 12:15):
We should ask Josh if he can come up with some queries?
nicola (RIO/SS) (May 12 2018 at 12:17):
Let's start from simple statistic - like patient, age ; age, visits etc
Tim Zallin (May 12 2018 at 13:16):
we ve just reuploaded new dataset into our postgres
Bryn Rhodes (May 12 2018 at 13:52):
@nicola (RIO/SS) , here's the github repository for the HEDIS-IG: https://github.com/cqframework/hedis-ig/tree/master/pages/cql
Bryn Rhodes (May 12 2018 at 13:53):
And here's the test data: https://github.com/cqframework/hedis-ig/tree/master/resources
nicola (RIO/SS) (May 12 2018 at 14:10):
@Josh Mandel supposed to implement just a one function for example - https://github.com/cqframework/hedis-ig/blob/master/pages/cql/ccs-cds-logic.cql#L100
nicola (RIO/SS) (May 12 2018 at 14:20):
@Bryn Rhodes we decided to pick this one - https://github.com/cqframework/hedis-ig/blob/master/pages/cql/ccs-cds-logic.cql. Is there open terminology server to get all valuesets for this script?
Bryn Rhodes (May 12 2018 at 15:12):
@nicola, yes: http://fhir.ext.apelon.com/dtsserverws/fhir
Bryn Rhodes (May 12 2018 at 15:13):
Basic authentication using (dtsadminuser/dtsadmin)
Bryn Rhodes (May 12 2018 at 15:15):
Looks like the CPT value sets aren't working, but the rest are coming back.
nicola (RIO/SS) (May 13 2018 at 07:43):
Thank you!
nicola (RIO/SS) (May 13 2018 at 12:01):
Here is experimental data set for cql - https://storage.googleapis.com/fhir-storage/cql.nd.json
nicola (RIO/SS) (May 13 2018 at 12:01):
It was generated for valueset "Cervical Cytology Value Set": '2.16.840.1.113883.3.464.1004.1208'
nicola (RIO/SS) (May 13 2018 at 12:28):
Here is how i could look in postgresql - https://github.com/fhir-fuel/fhir-storage-and-analytics-track/blob/master/postgresql/example.cql.sql
Tim Zallin (May 13 2018 at 12:38):
here is the link to report that we need to fill out in the end of todays session https://docs.google.com/document/d/1hrQt7pp4l7Mkk-jIOHTLI7cxkybnffo7k1NKDWa67RA/edit
Emmanuel Helm (May 13 2018 at 12:44):
I added myself to the participants. It was very interesting to see the storage model and query capabilities, but I failed to come up with meaningful data mining.
Tim Zallin (May 13 2018 at 13:46):
@Emmanuel Helm great thanks !
Last updated: Apr 12 2022 at 19:14 UTC
