GE HealthCare and Amazon Web Services (AWS) have formed a strategic collaboration to develop foundation models and generative AI applications.
GE HealthCare has tapped AWS as its strategic cloud provider and said it plans to use the firm's healthcare and generative AI services to build and implement foundation models. The companies said they will direct these generative AI workflows toward streamlining healthcare operations, increasing diagnostic and screening accuracy, improving patient outcomes, lowering access hurdles, and promoting health equity.
Making use of Amazon's Bedrock managed service, GE HealthCare plans to focus on developing multi-modal foundation models designed to analyze unstructured medical data, including images, records, and reports. The goal is to provide adaptable insights for healthcare applications, according to the vendor. The company's internal developers will also make use of Amazon Q Developer, a generative AI-powered software development assistant. What's more, it plans to utilize Amazon Q Business to explore the intersection of multimodal clinical and operational data, in hopes of reducing cognitive burden on physicians, facilitating personalized care, and increasing efficiency, GE HealthCare said.
In addition, the vendor said it expects to modernize its application suite with its own foundation models developed using Amazon's SageMaker managed service for developing and implementing machine-learning models. These generative AI-based applications will integrate with Amazon's AWS HealthLake and AWS HealthImaging services.
GE noted that it recently used foundational modeling for SonoSam Track, an advanced ultrasound image segmentation tool. The software achieved an accuracy of over 90% in isolating and identifying anatomical structures with little human oversight, according to the vendor.
![A normal mammogram confirmed by three-year radiologic follow-up illustrates reader-marked regions of interest (ROIs) during (A) unaided (round 1) and (B) artificial intelligence (AI)–assisted (round 2) reading. Each colored dot represents an ROI for recall by a human reader. Readers could mark more than one ROI per case, represented by multiple dots of the same color. During AI-assisted reading, the AI system displayed three visible prompts: two with suspicion of malignancy scores of 35% (left mediolateral oblique [L MLO] and craniocaudal [L CC]) and one with a suspicion of malignancy score of 10% (right craniocaudal [R CC]), shown as polygonal overlays. Without AI, six of 10 readers (60%) marked a false-positive ROI. With AI assistance, this fell to two of 10 (20%). R MLO = right mediolateral oblique.](https://img.auntminnie.com/mindful/smg/workspaces/default/uploads/2026/07/2026-07-14-radiology-mammogram-ai-auto-bias.H0bYO8QlWs.jpg?auto=format%2Ccompress&dpr=2&fit=crop&h=167&q=70&w=250)
