{"id":3242,"date":"2025-03-03T06:55:00","date_gmt":"2025-03-03T14:55:00","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/innovation\/blog\/ms-industry\/leading-the-charge-to-transform-healthcare-with-advanced-ai\/"},"modified":"2026-04-09T08:59:50","modified_gmt":"2026-04-09T15:59:50","slug":"leading-the-charge-to-transform-healthcare-with-advanced-ai","status":"publish","type":"ms-industry","link":"https:\/\/www.microsoft.com\/en-us\/microsoft-cloud\/blog\/healthcare\/2025\/03\/03\/leading-the-charge-to-transform-healthcare-with-advanced-ai\/","title":{"rendered":"Leading the charge to transform healthcare with advanced AI\u00a0"},"content":{"rendered":"\n

In today’s rapidly evolving healthcare landscape, AI is revolutionizing patient care by enabling more personalized experiences, optimizing vast medical data management, and improving patient outcomes. As challenges such as rising patient expectations, complex data handling, and regulatory requirements intensify, more advanced solutions have become essential. <\/p>\n\n\n\n

Microsoft is at the forefront of this transformation, dedicated to developing and implementing responsible AI technologies. By fostering innovation and collaboration through Microsoft Cloud for Healthcare<\/a>, we continue to reinforce how responsible AI<\/a> can enhance healthcare delivery and improve outcomes for patients worldwide. Building on this commitment, we\u2019re excited to introduce new features in our AI healthcare portfolio that will further drive industry efficiencies, and better patient outcomes. <\/p>\n\n\n\n

\n
Explore Microsoft Cloud for Healthcare<\/a><\/div>\n<\/div>\n\n\n\n

Advanced AI models and integrations for healthcare <\/h2>\n\n\n\n

As medical technology advances, improvements in medical imaging are critical for better diagnosis of disease and improved patient care. In 2024, we announced the launch of healthcare AI models<\/a>, a collection of cutting-edge multimodal medical imaging foundation models available in Azure AI Foundry. Designed for precise image segmentation, MedImageParse 2D model covers many imaging modalities, including x-rays, CTs, MRIs, ultrasounds, dermatology images, and pathology slides. It can be fine-tuned for specific applications such as tumor segmentation or organ delineation, allowing developers to test and validate the ability to leverage AI for highly targeted cancer and other disease detection, diagnostics, and treatment planning.  <\/p>\n\n\n\n

Today, we\u2019re excited to share the MedImageParse model is now optimized for 3D medical imaging data. MedImageParse 3D<\/strong> can handle complex 3D datasets produced by advanced imaging, such as MRI and CT scans, providing a more comprehensive view into patients\u2019 conditions. The enhanced ability to visualize and interpret anatomical abnormalities and structures provides for much more accurate diagnosis that may have been missed by 2D analysis. MedImageParse can also support healthcare researchers with comprehensive image analysis and a more streamlined workflow for radiologists, improving overall efficiency and reducing human error. MedImageParse 3D can soon be found in the Azure AI Foundry model catalog<\/a>.  <\/p>\n\n\n\n

In partnership with Microsoft Research, the Microsoft Health and Life Sciences model catalog will also feature several new and updated multimodal medical foundation models including TamGen for protein design, Hist-ai for pathology, and ECG-FM for electrocardiogram (ECG) analysis. <\/p>\n\n\n\n

Leveraging multimodal AI for improved health insights <\/h2>\n\n\n\n

Today, we are excited to announce new functionality in healthcare data solutions that allows customers to orchestrate multimodal AI insights <\/strong>directly into Microsoft Fabric<\/a>. Now in public preview, orchestrating multiple modalities (e.g., text, image, audio, video, and other forms of sensory input) of health data within Fabric allows healthcare organizations to generate a robust set of insights that help faster decision-making and improved patient outcomes. <\/p>\n\n\n\n

Customers can leverage Fabric to orchestrate multimodal AI insights by connecting their healthcare data to a variety of AI services and models. These AI-generated insights are then integrated back into the healthcare data estate to enable various use cases like creating targeted outreach and care plans by enriching clinical conversations with social determinants of health (SDOH) and sentiments. Another possible scenario is deriving quick insights and disease progression trends for clinical research by creating image segmentations and combining it with imaging metadata through Microsoft Power BI<\/a> reports. <\/p>\n\n\n\n

The orchestration capability includes five out-of-the-box examples to help customers connect and integrate to AI models: <\/p>\n\n\n\n

    \n
  1. Text analytics<\/strong> for health<\/strong> in Azure AI Language to extract medical entities from unstructured data such as diagnoses and medications, and the relations between entities.\u00a0\u00a0<\/li>\n<\/ol>\n\n\n\n
      \n
    1. MedImageInsight<\/strong> AI model<\/strong> in Azure AI Foundry to generate medical image embeddings from imaging data.  <\/li>\n<\/ol>\n\n\n\n
        \n
      1. MedImageParse<\/strong> AI model<\/strong> in Azure AI Foundry enables segmentation, detection, and recognition from imaging data across numerous object types and imaging modalities.  <\/li>\n<\/ol>\n\n\n\n
          \n
        1. Sentiment analysis<\/strong> with Azure OpenAI Service to score sentiment for categories such as doctors\u2019 services, staff services, facilities, and cost from conversational data. <\/li>\n<\/ol>\n\n\n\n
            \n
          1. SDOH <\/strong>e<\/strong>xtraction <\/strong>with Azure OpenAI to extract social determinants of health data from conversational data based on the Centers for Medicare and Medicaid Services\u2019 defined categories. <\/li>\n<\/ol>\n\n\n\n

            To further enhance data accessibility, we\u2019re pleased to share the general availability of additional functionality that enhances the existing capabilities within our healthcare data solutions offering. These include:   <\/p>\n\n\n\n