Why Health Care Firms Are Experimenting With Google's Generative AI
Google says its new proprietary A.I. could one day help accelerate drug discovery processes and make it easier to search through medical records.
BY BEN SHERRY, STAFF REPORTER@BENLUCASSHERRY
Google is betting that its generative A.I. can bring new efficiencies to the health care industry.
In April, the company made Med-PaLM 2, its large language model specifically designed to provide answers to medical questions, available to a select number of clients to explore use cases. In a recent blog post, the search giant stated that the clients are using the A.I. to efficiently generate notes, accelerate drug discovery processes, and make it easier to search through medical records.
A consistent pain point for medical practitioners is the overwhelming level of digital paperwork they have to generate, analyze, and maintain. In a 2021 study by the scientific journal Healthcare, more than 50 percent of health care providers said that they regularly experience stress related to the use of health care information tech. To help solve this issue, Google lent its PaLM 2 model to the medical documentation company Augmedix, which used the tech to create Augmedix Go, an application that allows doctors and nurses to generate medical notes just by recording conversations with patients.
The first health care provider to launch a pilot program for the A.I.-powered note-taking app is HCA Healthcare, which operates over 180 hospitals and many more clinics and care centers across the country. In the pilot program, approximately 75 emergency room physicians across four of HCA's hospitals started using Augmedix Go to create medical notes. The physicians then review the note to ensure there aren't any errors before transferring it to the hospital's electronic health record (EHR).
Augmedix CEO Manny Krakaris says his team has received positive feedback from medical professionals who have been seeking a better way to digitally take notes. "If you give a bottle of water to someone who's been in the Sahara Desert for a week without water, they're really gonna drink it up," Krakaris says. HCA is planning to roll out the program to more hospitals later this year.
Because EHRs can be incredibly detailed and complicated to sift through, Meditech, one of the largest EHR software providers, is exploring ways to use Med-PaLM 2 as a more powerful search engine for medical records. For example, Google wrote in the blog post that a clinician could ask a chatbot questions regarding a patient's condition, and the chatbot could surface key information like patient records, clinical guidelines, and research articles.
One of the costliest and most time-consuming projects in the health care industry is the development of new drugs and medications. The entire process can take anywhere from 12 to 15 years and cost more than $1 billion. Bayer Pharma, the German company best known for creating aspirin, was one of the companies given early access to Google's health care-focused A.I. The blog post stated that Bayer is using the tech to save time by automatically highlighting relevant connections between similar pieces of research data. The tool is also being used to both write and translate documentation related to clinical trials.
Google announced in the blog post that it will slowly open up accessibility to Med-PaLM 2 in September, but only as a preview, as it continues to build out the technology for a future full release.
Photo credit: Getty Images.