Medical Entity Abstraction for a Real-World Data Provider

  • Published: March 28, 2024



Understand how a real-world data provider was able to increase efficiency and reduce costs of their operations on Reveal’s recommendation and advice.

So Many Medical Records, So Little Time

Our client was finding it increasingly challenging to manage unstructured medical records, particularly due to manual processes. Extracting clinical insights from this data consumed considerable time. Moreover, the ongoing expansion of medical records underscored the pressing need for efficient techniques to extract valuable information from them.

Bringing Method to the Madness

Once Reveal understood the scope of the challenge, they built a proof-of-concept model for medical entity extraction. It was developed using SparkNLP and other open-source natural language processing (NLP) models. Reveal also evaluated and compared the performance of multiple third-party commercial NLP services for consideration as potential vendors.


Building Trust & Efficiency

Following the POC done by Reveal, our client successfully extracted valuable insights from their unstructured data. They observed significant improvements in efficiency and operational cost reduction. Consequently, the client has chosen to continue their partnership with the vendor recommended by Reveal.