Thank you @d.vanstory01 for kicking off this project! That’s very cool.
A few notes/considerations:
- The Bahmni distribution of OpenMRS has done something similar using Medispeak: https://bahmni.atlassian.net/wiki/spaces/BAH/pages/3946381343/Medispeak+Integration+Guide+Streamlining+Bahmni+EMR+Data+Entry+with+Voice+Assistance I’ve often wondered what it would take for us to add Medispeak or something similar to the O3 EMR/RefApp, and I’ve recently heard good things about Whisper, so very curious to hear the ongoing progress of your work
Video:
- Please keep in mind that in order for the solution to be used in production settings, Patient data must not be sent to an outside service, such as a SaaS offering like ChatGPT’s SaaS offering, as this violates many country’s health data privacy laws and/or may breach patient consent. And also there’s the practical limitation of internet access in most OpenMRS sites; so a solution that can run performantly on a local machine (on-prem or laptop) would be really excellent!
Of course using a web API call to a GPT service just for demo / test purposes is still very interesting; just wanted to share these 2 limitations (privacy and internet) for our on-site settings.