In a recent blog post, Google Cloud’s global director for healthcare strategy and solutions Aashima Gupta and head of health AI Greg Corrado highlight how three early adopters have used Google’s generative AI Med-PaLM 2 to support their organisations, through standardising processes, supporting creation of medical documentation, and more.
Med-PaLM is a “large language model designed for the medical domain” capable of providing “high quality answers to medical questions”. The latest model, Med-PaLM 2, was launched in March this year. It is currently available to a “select group” of customers to enable Google to examine use cases and feedback, ahead of it becoming available to more customers in the healthcare and life sciences industry next month.
The blog shares how Bayer Pharmaceuticals is trialling generative AI solutions such as Med-PaLM 2 and Google Cloud’s Vertex AI to see how they can assist in bringing drugs to market. Google notes that generative AI can support researchers in this area by enabling them to “more easily access, identify and correlate data, mine large troves of research data for possible connections” along with automating time-intensive tasks such as drafting communications around clinical trials and translating them into different languages.
Another early adopter, HCA Healthcare, has also been using generative AI to tackle the burden of administrative tasks, including piloting a solution that supports the creation of medical notes by extracting information from physician-patient conversations. The solution sees providers using an app built by Augmedix that “securely creates draft clinical notes automatically after each patient visit”, which physicians can then review and finalise before they are transferred to the electronic health record in real time. HCA Healthcare is also exploring ways to use generative AU to standardise and automate the process of patient handoffs between nurses.
MEDITECH, meanwhile, is currently using Google’s AI to power their electronic health record search and summarisation functions, with the aim of using the technology to combine information from different sources to create a longitudinal view of the patient’s record. In addition, MEDITECH is exploring ways in which Med-PaLM 2 can “help their clinicians gain “a deeper understanding of a patient’s history”, and how AI can help with clinical documentation such as discharge summaries, to “improve the efficiency of care delivery”.
Earlier this month, we looked at Google Health’s work to build “lightweight, multimodal generative AI models for medical imaging” which use a combination of large language models (LLMs) and vision encoders for X-rays.
We also covered Microsoft and Epic’s announcement that they are expanding their strategic initiative combining Microsoft’s large-scale cloud and artificial intelligence technologies with Epic’s electronic health record ecosystem and healthcare industry experience, with the aim of addressing current issues affecting clinicians such as workforce burnout and staffing shortages.
