Posted by: RedaChouffani
EHR, information extraction, NLP, text analysis, text parsing
An area of computer science long studied by university graduates and professors is becoming one of the leading informational technologies: natural language processing.
Many markets already implement NLP. Apple uses it in accordance with Siri. Nuance uses a health care specific model offered alongside clinical language understanding. Certain search engines use it to translate user queries into structured responses.
Thinking long term, NLP would provide nearly endless possibilities in the way we interact with computer systems. By leveraging the increase we have seen in both computing powers and storage capability, it could have a great impact on at least two major areas of healthcare.
Text parsing and information extraction:
The use of NLP in regard to text parsing is focused on the extraction of information from unstructured or semi-structured documents like text, dictations or scanned and indexed health records. This information is then used either for reporting or collection and indexing for future use. These text parsing tools help clinicians extract meaningful, actionable information from data stored in unstructured formats, whether those are EHRs or other health information sources.
Natural language to command or sort queries:
NLP interacts with users through voice, which may seem to be an idea torn from a sci-fi flick. But consider examples such as IBM’s Watson, or some of the pilot projects previously made around natural language interfaces to databases. We can ask simple questions and a computer system can translate them into actual commands or queries that can in turn be used to look up specific information.
In a clinical setting, we are seeing basic integration capabilities in several voice recognition products that not only allow for voice commands, but also discrete data capture through voice. And as the field continues to advance, these systems will mature and enable us to interact with information directly by using our voice and commands. This will allow clinicians to not only completely avoid typing to capture information, but also to simply ask for information about their patients and have that data displayed on a screen. Theoretically, they might never need a keyboard.
NLP is right alongside big data initiatives, IVR systems, search engines, sentiment analysis and several other areas that continue to see a significant growth in invention and adoption. That being said, there are still only a handful of examples within healthcare that showcase how NLP is changing the way we interact with information.