Google Unveils MedLM Healthcare Generative AI Models
Google has introduced a suite of foundational generative AI models tailored for the healthcare industry called MedLM. The new large language models are available to Google Cloud customers for both providing healthcare and medical research.
MedLM evolved out of Google’s work on the Med-PaLM model and its successor. The new model enhances the features of the earlier medically-tuned LLMs and offers different flavors depending on the needs and goals of the healthcare organizations employing it. There are two distinct MedLM options and many ways to tailor them beyond that point, though both are capable of answering medical queries and drafting healthcare summaries. The progress seen with MedLM already has several notable achievements, including passing the U.S. medical licensing exam and to applying it in real-world scenarios.
“The first MedLM model is larger, designed for complex tasks. The second is a medium model, able to be fine-tuned and best for scaling across tasks. The development of these models has been informed by specific healthcare and life sciences customer needs, such as answering a healthcare provider’s medical questions and drafting summaries,” Google vice president of engineering and research Yossi Matias and Google Cloud global director of healthcare strategy and solutions Aashima Gupta explained in a blog post. “Many of the companies we’ve been testing MedLM with are now moving it into production in their solutions, or broadening their testing.”
The official debut of MedLM will enable Google to bring the LLM and its specialized intelligence across healthcare systems and life sciences companies. The potential and current use cases include answering doctors’ questions, summarizing patient data, and analyzing diagnostic tests for both current patients and medical studies. Google said strict protocols govern MedLM’s deployment to ensure responsible and safe AI adoption in the heavily regulated healthcare industry. The company envisions the technology significantly improving patient outcomes and operator efficiency over time.
Even before MedLM, there’s been enthusiasm for the kind of service it provides. For instance, the Mayo Clinic piloted integrating Google’s Med-PaLM 2 into its research hospital as part of its broader partnership with Google Cloud. Meanwhile, Microsoft subsidiary Nuance started incorporating GPT-4 into the Dragon Ambient Intelligence platform, Amazon released the AWS HealthScribe, and the British National Health Service is pushing hospitals to adopt generative AI tools with $27 million in grants. And there are already ongoing projects in pharmaceutical research of drugs conceived and developed using generative AI underway.