Healthcare Natural Language API and the AutoML Entity Extraction for Healthcare are two new artificial intelligence tools designed to help healthcare and life science organisations scan and analyse large volumes of unstructured text.
The former automatically extracts information from medical record notes or other digital texts that would normally require time-consuming manual review. Healthcare Natural Language API identifies medical concepts in documents, automatically extracting knowledge about medical procedures, medication, body vital signs or medical conditions. Using machine learning, the API identifies clinically relevant attributes based on the surrounding context. For example, it can discern medications prescribed in the past from those prescribed for the future and detect the likelihood of a specific symptom or diagnosis, as expressed in the nuances of language. It can also distinguish medical concepts that concern the patient from information that concerns his or her relatives.
The Healthcare Natural Language API can be used in a variety of medical fields including televisit applications to store transcribed conversations between the doctor and a patient, as well as clinical trials that recruit patients according to specific inclusion or exclusion criteria.
In addition to the Healthcare Natural Language API, Google has launched AutoML Entity Extraction for Healthcare – an easy-to-use AI development platform that extends access to AI to all users with different technical backgrounds.
Healthcare professionals may not have the technical skills to build their own tools for extracting information from digital documents. AutoML Entity Extraction for Healthcare makes this much easier with an interface that requires little code writing, allowing healthcare professionals to build information extraction tools for gene mutations and socioeconomic factors, for example.
Both tools are now available in preview. The Healthcare Natural Language API is available free for businesses until 10 December, while the AutoML Entity Extraction for Healthcare is free for the first 5,000 text records and 1,000 pages of imported documents.
Digital-health 