NLP Helps Extract Key Information from Patient Records
March 24, 2021, 2 Minute Read
Recently, we were engaged on a proof of concept to “speed read” patient records, comprised of their textual information such as physicals, labs, and histories.
The extractions were performed on example data collected from a Provider’s electronic medical record system.
We used the power of our Lymba NLP Platform to understand information in three categories:
Understanding Clinical Findings (symptoms and diagnoses of patients)
Determining Clinical Procedures (collecting the tests, skills applied, drugs, and devices used on patients)
Extracting any specific Drugs used (collecting the drugs, brands, dose form, strength prescribed for patients)
We were asked to apply Ontologies (or concept classifications) to help understand, identify, and prepare insights surrounding these categories.
We found that, first, with our platform, it has become incredibly accessible to generate quick wins from this data. Instead of a human reading through charts, our data can be directly populated to knowledge graphs for immediate analytics. Imagine having this data at your fingertips to understand clinical efficiencies, improving clinical guideline adherence, or even cross-checking recommendations for drug sensitivities or clinical interventions.
As next steps, we also found that implementation of our solution to prepare for ICD-11 compliance for medical billing and coding is a perfect fit. The harmonized agreements to implement ICD-11 over the next couple of years is ideal for knowledge graphs and NLP to show how these technologies can improve patient care. We can help demonstrate quick wins with the Lymba NLP Platform today – feel free to reach out for a consultation and demo.
Schedule your free consultation below: