Dive Brief:
- Google Cloud unveiled generative AI capabilities aimed at solving workforce shortages, burnout and administrative burden in the healthcare and life sciences industries on Monday.
- The company plans to add medically tuned, generative AI-powered search capabilities that connect to clinical notes, data from the industry standard Fast Healthcare Interoperability Resources protocol and other clinical sources within Vertex AI, a platform that lets clients train and deploy AI models.
- The search capabilities work alongside Google’s healthcare-focused large language model, Med-PaLM 2, connecting to outside sources and the patient’s medical record. Customers can sign up for early access starting Monday.
Dive Insight:
In an effort to win over healthcare industry clients and capture growth, cloud providers are releasing specialized tools to help mitigate their pain points. While risk-averse and heavily regulated, the industry also increasingly depends on tech solutions.
Aashima Gupta, global director of healthcare for Google Cloud, declined to share how many systems are using Vertex AI Search at the HLTH conference in Las Vegas, where Google made the announcement.
But Google is rolling out the capabilities with Minnesota academic medical system Mayo Clinic, Pittsburgh system Highmark Health and New Jersey-based Hackensack Meridian Health, according to a release from the company.
Mayo Clinic plans to use Google Cloud’s generative AI search capabilities in Vertex AI to use data to support a wide range of applications, said Vish Anantraman, chief technology officer for Mayo Clinic, in the announcement.
Other Google Cloud clients can also sign up for early access to Vertex AI, Gupta said.
Microsoft has also released generative AI tools with the goal of helping healthcare companies assist patients, improve scheduling procedures and power an administrative chatbot, according to the company. The cloud giant entered a long-term partnership with Mercy in September to deploy Azure OpenAI Service tools, exploring more than four dozen use cases.
Oracle has joined in, aiming to tailor its services to healthcare companies. The company’s “vector database will contain highly specialized training data, like electronic health records, while keeping that data anonymized and private,” Oracle CTO and Chairman Larry Ellison said during a September earnings call for Q1 2024. The tool can help doctors improve diagnostic capabilities and treatment prescriptions, Ellison said.
Even if generative AI presents opportunities for healthcare workers and systems to become more efficient, solutions must meet rigorous security requirements. Technology leaders in the space are tracking internal governance models as well as those of the vendors they choose to partner with.
Google Cloud’s question-and-answer conversational search applications follow the same security protocols as other tools that the company has released for the healthcare industry.
Customers retain control over their data with access and use of patient data protected through Google Cloud’s infrastructure and secure data storage, which supports HIPAA compliance.
During a fireside chat on Monday, James Manyika, Google’s senior vice president of research, technology and society, said the tech giant has learned from past missteps in healthcare, including accusations from British regulators in 2017 that AI division DeepMind was breaking data protection laws and a controversial data sharing and product development partnership with Ascension in 2019.
”We’ve learned a lot from all of these things,” Manyika said, but “as the technology develops,” the industry will face new questions about data privacy.
During a Google media briefing at HLTH, members of its health team said Google is working closely with the government, academia and other stakeholders to create standards around AI usage in healthcare.
Read More: Google Cloud tunes generative AI for medical search capabilities