Google Cloud pulled back the curtain yesterday on two artificial intelligence tools designed to help healthcare and life science organizations scan and analyze large volumes of unstructured text, the Healthcare Natural Language API and AutoML Entity Extraction for Healthcare.
The first of these two offerings looks to automatically extract common trends or other insights from medical records notes or other digital text that would normally require time-intensive manual review. According to the company, the machine learning tool discerns clinically relevant information based on the context of surrounding language, allowing the technology to, for instance, distinguish between past and newly prescribed medications.
The Healthcare Natural Language API can be deployed within a provider organization for analysis, or can be implemented within a range of health applications that support unstructured text, Google said. Potential use cases supplied by the company in a blog post include a telehealth app that stores transcribed conversations between the doctor and a patient, as well as clinical trials enrolling patients based on specific inclusion or exclusion criteria,
AutoML Entity Extraction for Healthcare, meanwhile, seeks to lower the barrier to AI text data analysis for healthcare workers. According to the company, it provides an easier-to-use interface that helps less experienced users train their own machine learning analysis models. It has, for example, a tool that extracts information on patients’ relevant gene mutations, or on socioeconomic factors.
Google launched both tools yesterday in public preview. The Healthcare Natural Language API is available to enterprises for free until Dec. 10, while AutoML Entity Extraction for Healthcare is free for the first 5,000 text records and 1,000 document pages imported.
WHY IT MATTERS
Unstructured data housed within EHRs or other medical notations house a wealth of relevant patient data that could be used for clinical research, accurate clinical modeling or streamlining other administrative tasks. However, the large volume of data organizations generate each day would require a substantial effort to record and analyze by hand.
“For healthcare professionals, the process of reviewing and writing medical documents is incredibly labor-intensive,” Andreea Bodnari, a product manager at Google Cloud, wrote in the blog post announcement. “And the lack of intelligent, easy-to-use tools to assist with the unique requirements of medical documentation creates data capturing errors, a diminished patient-doctor experience, and physician burnout.”
Google Cloud’s tools are the latest seeking to address the longstanding challenge of unstructured data, and notably seek to do so with guided interfaces and other considerations that make these analyses more accessible to healthcare staff.
THE LARGER TREND
Google isn’t the only major tech player that’s aware of natural language processing’s potential in healthcare. Its chief rival, Amazon Web Services, launched its own tool called Amazon Comprehend Medical near the tail-end of 2018, and similarly highlighted the ease with which the machine learning tool could be deployed within an enterprise’s existing systems.
There’s certainly interest among healthcare organizations for this type of technology. Just this week, Centene announced plans to purchase the patient data analytics platform Apixio in order to deploy similar text-analysis technology across its managed care enterprise.
More regional news
Suki adds six to its leadership team, former Cleveland Clinic president and CEO joins AliveCor's board and more digital health hires
Ohio rolls out service to match patients with beds in psychiatric facilities
Practicing 'neurobics' before surgery found to reduce risk for postoperative delirium
Source: Read Full Article