natural language processing solutions

Makes it easy to find, use and unlock the potential of your data

Natural Language Processing applies machine-learning techniques to human language to generate meaning and structure from vast sources of textual information.

A significant proportion of the world’s health information is in the form of “unstructured text” – notes, letters, summaries and other types of written information. Often, this information isn’t coded or tagged, and is therefore difficult and time-consuming to use, or remains essentially invisible to systems and processes that could generate significant value from it. 

Orion Health’s suite of Natural Language Processing solutions includes the Smart Coder for automatic coding, and the ability to develop custom models from your data. 


Why you will love our tech

Make it easy for clinicians to find and use the information they need, when they need it
Enhance safety by minimizing the risk of missed information or errors occurring
Improve organizational data quality
Enable new applications such as reporting and analytics
The structured data generated by applying Natural Language Processing unlocks the potential of machine learning

Value our tech delivers

Automatic detection and tagging of clinical codes

  • Identify clinically relevant phrases from free text, automatically. Named entities include disorders, procedures, medications and findings. 
  • Accurately process the context of clinical concepts such as negation. For example, “no family history of heart disease” would be detected as a negated concept. 

Sophisticated applications such as cohort-finding

  • Advanced uses of Natural Language Processing solutions can be applied to find cohorts of patients where information about them is otherwise hidden in textual information. 
  • This benefits population health management, personalized care and life sciences applications.

Generates structured, usable data from unstructured text

  • The processed, coded data that is output can be easily used by other applications to search for information or to generate analytics and reporting. 

Linking to code sets including SNOMED and ICD-10

  • Deep learning-based approaches normalize free text to clinical ontologies such as SNOMED CT and ICD-10.
Find out more
Reimagine the healthcare experience for all