Microsoft will launch medical imaging foundation models as the first release of its new healthcare AI models in Azure AI Studio, which will allow healthcare organisations to test, fine-tune and build AI solutions tailored to their specific needs.
Part of some major innovations added to the Microsoft Cloud for Healthcare announced this week in advance of the HLTH conference, Microsoft is also planning to introduce conversational data integration which will allow clinicians to send conversational data such as patient conversations from DAX Copilot to Microsoft’s Fabric platform.
It is also adding a healthcare agent service in Copilot Studio, and an AI-driven nursing workflow solution.
Microsoft’s corporate vice president for healthcare and life sciences solutions and platforms Joe Petro said the world was at an inflection point where AI breakthroughs are fundamentally changing the way we work and live.
“Across the broader healthcare and life sciences industry, these advancements are dramatically enhancing patient care and also rekindling the joy of practising medicine for clinicians.”
Developed in collaboration with Microsoft Research and partners, the new AI models are designed for healthcare organisations and their specific needs, promising to minimise the extensive compute and data requirements typically associated with building multimodal models from scratch.
The models include MedImageInsight, an embedding model that enables sophisticated image analysis, including classification and similarity search in medical imaging. These can be used in radiology, pathology, ophthalmology, dermatology, and other modalities.
Microsoft gave the example of researchers who can explore how the model can be used to build tools to automatically route imaging scans to specialists, or flag potential abnormalities for further review.
MedImageParse is designed for precise image segmentation for x-rays, CTs, MRIs, ultrasounds, dermatology images, and pathology slides, and CXRReportGen, which can generate detailed, structured reports from chest x-rays, highlighting AI-generated findings directly on the images to align with human-in-the-loop workflows.
Microsoft has also made healthcare data solutions generally available on its Microsoft Fabric platform, a solution that allows organisations to ingest, store and analyse healthcare-related data from various sources and modalities into one unified data store for analytics and AI.
In public preview in Microsoft Fabric are conversational data integration, which will allow patient conversations, audio files, transcripts and draft clinical notes to be sent from DAX Copilot to the Fabric platform.
Also in public preview is its healthcare agent service in Copilot Studio, which will allow users to build Copilot agents for appointment scheduling, clinical trial matching and patient triaging. This is being adopted by medical centre Cleveland Clinic.
The new solutions are being released at the HLTH 2024 conference in Las Vegas next week.
