Microsoft announced Thursday that it is unveiling several artificial intelligence enhancements in Microsoft Cloud for Healthcare innovations, including new healthcare AI models in Azure AI Studio, new healthcare data capabilities in Microsoft Fabric, and developer tools in Copilot Studio. Many of the artificial intelligence improvements are available in preview.
Company executives also provided additional details about the AI-powered nursing workflow partnership with Epic during a media briefing on Tuesday.
“Across the broader healthcare and life sciences industry, these developments dramatically improve patient care and also reignite the joy of practicing medicine for physicians,” said Joe Petro, corporate vice president of healthcare and life sciences solutions and platforms at Microsoft, in a statement explaining the new possibilities.
Capturing GenAI data for clinical insights
The ability to integrate structured and unstructured data into Microsoft Fabric will help reshape the way users access, manage and interact with data, the company said.
Some of the new AI-driven innovations can combine data from electronic health records to generate comprehensive insights and drive use cases including clinical imaging, Centers for Medicare and Medicaid Services claims, social determinants of health and more, said Umesh Rustogi, general director manager of Microsoft Cloud for healthcare.
“We’ve had organizations around the world that have actually started working on the solution to create a unified data hub that allows them to create not only new insights, but new AI models to improve patient care and create outpatient efficiencies” , Rustogi said during the briefing.
An important new feature is the integration of conversation data.
The generative AI voice-driven tool, Nuance’s DAX Copilot, has been generally available for a year, and the company noticed in a blog post there has been remarkable momentum over the past month.
It helps doctors at Northwestern Medicine in Chicago make at least 50% of patient encounters, reducing their average time on writing notes by 24% and increasing the number of patients they can see by an average of 11.3, the company said.
The healthcare system integrated the AI copilot with its Epic EHR.
“This integration now allows organizations to securely access the DAX Copilot call data,” including the audio files, draft clinical notes and more, Rustogi said.
Using proprietary analytics tools in Azure and Fabric, healthcare organizations can analyze the data and combine it with other patient data, such as EHR data and patient engagement insights, to create comprehensive data.
For example, by consuming and harmonizing national and international public SDOH datasets, healthcare organizations can identify risks and health-related social needs to improve healthcare equity.
The tools can also leverage unified healthcare data and care management analytic templates to improve patient care by identifying high-risk individuals, optimizing treatment plans and improving care coordination, the company said.
New templates for healthcare security applications that can help manage data are also available in public preview in Microsoft Purview, the company said.
Verify GenAI output
A healthcare-specific stack in Copilot Studio – including pre-built healthcare information and use cases – is now securely accessible, Hadas Bitran, head of health and life sciences at the Microsoft Israel R&D Center, said on Tuesday.
Bitran leads the development of AI-powered language services, natural language processing technologies, conversation intelligence and personal health assistants. For developers, the Clinical Safety Measures API is available in private preview for additional use cases, she said.
“For example, these APIs can be used for evaluation and additional verification of the generative AI output,” she explains. “Clinical familiarity helps identify the source of claims in the answers against the basic data or facts.”
The assurance of clinical semantic validation helps examine whether responses meet verified clinical standards and ensure trust and stability, she noted.
Automated documentation for nurses
Through a partnership with Epic Systems announced in August, Microsoft is building an AI-powered documentation tool for nurses that Mary Varghese Presti, vice president of portfolio evolution and incubation at Microsoft Health & Life Sciences, says could leverage the use of Epic Rover to to provide the best results. nursing-focused experience.
She explained in the media briefing that the inspiration for the nursing workflow collaboration – apart from the high levels of nurse burnout due to administrative burdens – is to enable nurses to be “eyes and hands free” in their documentation.
“We can transform nursing documentation through voice,” she said. “We are also acutely aware that collaboration results in the best experience for nurses. And our goal was to enable a seamless workflow – and not add any extra steps.”
The company said it is actively working with Advocate Health, Baptist Health of Northeast Florida, Duke Health, Intermountain Health Saint Joseph Hospital, Mercy, Northwestern Medicine, Stanford Health Care and Tampa General Hospital to develop environmental technology that handles nursing documentation by creating flowcharts proposals for assessment.
“AI is transforming nursing workflows by streamlining administrative tasks, allowing nurses to focus more on patient care,” said Corey Miller, vice president of research and development at Epic, in a statement.
“Together with Microsoft, we are using AI-powered ambient voice technology to populate patient assessments. Nurses using the tool are already sharing positive feedback about how it improves personalized patient interactions.”
Streamlining AI development
A collection of basic multimodal medical imaging models available in the Azure AI Model Catalog analyze various data types, including genomics and clinical records.
They were developed in collaboration with partners such as Providence and Paige.ai, the company said. It allows healthcare organizations to quickly build, refine, and deploy custom AI tools while minimizing the compute and data requirements typically associated with building multimodal models from scratch.
“The development of fundamental AI models in pathology and medical imaging is expected to deliver significant advances in cancer research and diagnostics,” said Dr. Carlo Bifulco, Chief Medical Officer of Providence Genomics, said in a statement.
“These models can complement human expertise by providing insights beyond traditional visual interpretation and will reshape the future of medicine as we move toward a more integrated, multimodal approach.”
Building secure AI agents
New tools in Microsoft Copilot Studio also allow healthcare systems to build custom AI agents for appointment scheduling, clinical trial matching, patient triage, connected patient experiences, improving clinical workflows and more.
Early adopters, such as Cleveland Clinic, helped develop the service by providing feedback on its use in a healthcare setting and are already using it, Microsoft said.
There is a public preview of the health care agent service available.
Andrea Fox is editor-in-chief of Healthcare IT News.
Email: afox@himss.org
Healthcare IT News is a HIMSS Media publication.