Google Research unveiled MedLM, a family of foundation models fine-tuned specifically for healthcare industry use cases. In a blogpost, Google says that it is is available on the Vertex AI platform, MedLM aims to enhance various healthcare applications and workflows.
MedLM Models for Diverse Healthcare Needs
MedLM comprises two models built on Med-PaLM 2, each tailored to address specific healthcare tasks. The larger model is designed for complex tasks, while the medium model offers flexibility and scalability across various applications. The development of these models has been shaped by the diverse needs of healthcare and life sciences customers, emphasizing the importance of a task-specific approach.
Real-World Implementations
Several companies have been testing and implementing MedLM in their solutions, showcasing its practical applications in healthcare.
Augmedix at HCA Healthcare
HCA Healthcare has successfully piloted Augmedix’s solution, leveraging MedLM to automate ambient medical documentation in emergency department hospital sites. Physicians use Augmedix’s platform, powered by MedLM on Vertex AI, to convert clinician-patient conversations into accurate medical notes in real time. This not only ensures compliance with privacy regulations but also enhances efficiency, reduces burnout, and improves patient care.
BenchSci’s ASCEND Platform
BenchSci integrates MedLM into its ASCEND platform, aiming to expedite pre-clinical research and development in drug discovery. ASCEND, an AI-powered evidence engine, decodes over 100 million experiments from various data sources, aiding scientists in understanding complex connections in biological research. The addition of MedLM enhances accuracy and reliability, propelling scientific discovery by identifying, classifying, and discovering novel biomarkers.
Accenture’s Solutions.AI for Processing for Health
Google Cloud collaborates with Accenture to empower healthcare organizations with generative AI, enhancing patient access, experience, and outcomes. Accenture’s Solutions.AI for Processing for Health interprets structured and unstructured data to automate manual processes such as reading clinical documents and claims processing. MedLM, in conjunction with Google Cloud’s Claims Acceleration Suite, uncovers valuable insights, ultimately leading to improved patient outcomes.
Deloitte’s Provider Search Enhancement
Deloitte and Google Cloud team up to explore how MedLM’s generative AI capabilities can improve the member experience in healthcare. By using interactive chatbots, MedLM assists health plan members in understanding provider options covered by their insurance plans. This collaboration aims to streamline administrative tasks, enhance member experience, and provide faster access to precise care.
As MedLM expands its reach to more healthcare organizations, the transformative potential of generative AI in healthcare becomes increasingly apparent. Google’s commitment to a safe and responsible use of technology involves close collaboration with practitioners, researchers, and healthcare organizations. From achieving expert-level scores in U.S. medical-licensing-exam-style questions to real-world implementations, MedLM marks a significant step forward in the integration of AI into the healthcare landscape. Looking ahead, the focus remains on pushing breakthrough research in health and life sciences, anticipating further progress and potential in the coming years.
