Apr 27 2021 09:07 AM
I'm interested in tools for parsing, understanding and ML from unstructured medical data. Is Azure Cognitive Services being used for this type of data?
Apr 27 2021 09:17 AM
@normkatz1, our Text Analytics for health feature extracts and labels relevant medical information from unstructured texts such as doctor's notes, discharge summaries, clinical documents, and electronic health records. You can learn more about it here: How to use Text Analytics for health - Azure Cognitive Services | Microsoft Docs
Apr 30 2021 10:03 PM
@KelvinChen Thanks for the tip. I'll definitely take a look.
If you work with Trent Norris, please tell him I said hi!
Cheers, Norm
May 20 2022 05:43 AM
@normkatz1 just saw this thread from your side, although it is almost an year old.
Curious to understand the use case you implemented using this feature of Health Text Analytics.
I am using the same but am finding that it does not work well with ICD10CM Data Source for billing code, which by the way is the most commonly Data Source used in the US.
Thanks.
May 20 2022 09:49 AM - edited May 20 2022 11:38 AM
@sheepoo I got pulled off that project and was not able to do a full evaluation. I started the project up again last September but we looked at AWS Comprehend Medical, and some higher end products. One of the NLP-specific products we used was clearly superior to the others in terms of parsing, classifying and codifying medical terms from unstructured data but the product is somewhat expensive and their learning curve is also the most complex among those that we looked at. If you're a Python developer with experience with building models for NLP, and a budget of $50K to $100K, then I can provide some more details about our eval and talk about how our company could help you and GE. Let's try to set up a time to connect.
May 21 2022 01:21 AM
@normkatz1 Thanks for the reply.
How would you compare MS Health Text Analytics with AWS Medical Comprehend? I have not had a chance to test AWC MC out yet, though.
Will see what is the best time for us to connect and discuss a little more.
Thanks
May 22 2022 01:52 AM