listen to article
This voice is automatically generated. Please let us know if you have any feedback.
Microsoft on Thursday announced a slew of new artificial intelligence capabilities for healthcare organizations, including a product that helps companies build their own AI agents.
The tech giant also announced the foundational model for its medical imaging and medical data analysis platform, as well as details of its plans to build an AI documentation product for nurses.
Healthcare organizations are increasingly interested in adopting AI tools, even as some experts and lawmakers question their safe and fair use. The tech companies say these products could help providers manage workloads and reduce burnout.
“We are at a tipping point. AI breakthroughs are transforming and extending the way we work and live,” said Vice President of Healthcare and Life Sciences Product Marketing. Kees Hertogh said at a press conference. “The integration of AI in healthcare is greatly enhancing patient care and rekindling the joy of practicing medicine for clinicians.”
Healthcare AI Agent
Microsoft's Agent service allows companies to create AI tools with pre-built templates and data sources for appointment scheduling, clinical trial matching, patient triage, and more, Hartog said. He says it can be done.
The service is currently in public preview, which will allow broad access to the tool and allow organizations to provide feedback on the product, a spokesperson said.
As an example, a doctor could ask an AI agent to find a clinical trial for a 55-year-old patient with diabetes and interstitial pneumonia.
This tool could also be aimed at patients. Cleveland Clinic used Agent Services to build tools that allow patients to ask health questions and navigate the health system's services, according to a spokesperson.
Microsoft's Copilot Studio for building AI agents
Provided by Microsoft
The product enables organizations to use intelligence from trusted sources to build agents with healthcare-specific capabilities, said Hadas Bitran, Partner General Manager, Health AI, Microsoft Health and Life Sciences. , states that the purpose is to improve safety.
The tech company will also release in private preview additional application programming interfaces (APIs) to help validate model outputs, including the ability to detect omissions and link answers to evidence data, he said.
imaging model
Microsoft also unveiled a system, or foundation model, built on a wide range of datasets that can be used for many tasks, with a focus on medical image processing.
The model was developed in collaboration with partners including health system Providence and digital pathology company Paige.ai, and healthcare organizations will need the extensive data and computing resources needed to build it from scratch. Now you can build your own AI tools without having to.
The underlying model includes MedImageInsight, which enables image analysis that can be used to automatically send scans to specialists and flag abnormalities for examination.
MedImageParse is intended for image segmentation. It can be used to segment tumors or outline organs at risk before radiotherapy in cancer patients.
The third model, CXRReportGen, creates a report based on a chest X-ray. Microsoft says this could speed up image analysis and improve diagnostic accuracy for radiologists.
Healthcare data analysis platform
The technology giant said healthcare-specific data tools are now generally available in its analytics product, Microsoft Fabric. This platform allows organizations to capture, store, and analyze health data.
“Analytics enhancements allow for enhanced reporting of patient geographic distribution, details such as age and gender, and patient outcomes and satisfaction status,” said Hertogh.
Organizations will also be able to use other data types in public preview, including conversation information from Microsoft's DAX Copilot AI document tool, public social determinants of health information, and claims data from CMS. Also in public preview, the tech giant released healthcare security templates for its data security and governance product, Microsoft Purview.
nursing documentation
Microsoft is working with electronic health records vendor Epic and several health systems, including Advocate Health, Northwestern Medicine, Duke Health, and Stanford Healthcare, to develop AI documentation tools for nurses.
Microsoft-owned Nuance Communications made its automated clinical documentation tool DAX Copilot generally available last year. This tool records conversations between clinicians and patients and creates clinical summaries.
According to Mary Varghese Presti, vice president of portfolio evolution and incubation at Microsoft Health and Life Sciences, the nurse product has been deployed at multiple customer sites. This tool uses ambient audio technology to automatically create flowsheets, or forms that collect patient data, for nurse review.
“For nurses, documentation is really data entry. They have their backs to the patient and their faces are at the computer,” Presti said. “(…) We want to enable nurses to document without using their eyes, without using their hands.”
