2024/8/30 11:43:27
Views:
GE Healthcare plans to create and deploy customized generative AI applications through Amazon Bedrock, further expanding the advantages that generative AI brings to customers. Amazon Bedrock is a fully managed service provided by AWS that allows users to easily access a variety of industry-leading large language models. Leveraging the enterprise-grade security, privacy features, and a rich selection of foundational models offered by Amazon Bedrock, GE Healthcare plans to develop and expand its proprietary generative AI applications to address various complex scenarios in healthcare, enhance efficiency, optimize care delivery, and significantly improve patient experience.
GE Healthcare's internal development team is preparing to use Amazon Q Developer, an AI-powered generative programming assistant that provides real-time code suggestions to help developers build applications faster and more securely. Additionally, GE Healthcare plans to utilize Amazon Q Business to deeply analyze the multidimensional intersections of clinical and operational data, reducing the workload of professionals, thus advancing personalized healthcare services and enhancing overall operational efficiency. By leveraging AWS's advanced technologies, GE Healthcare expects to significantly shorten development cycles and accelerate the implementation of innovative solutions.
"We are ushering in a new era in healthcare by combining generative AI with our extensive industry experience," said Dr. Taha Kass-Hout, Chief Technology Officer of GE Healthcare Global. "Our partnership with AWS marks an important step in simplifying medical processes, enhancing efficiency, and delivering deeply personalized services. Each innovative solution we introduce is designed to gradually improve the quality of care on a global scale."
Moreover, GE Healthcare plans to improve its application suite by creating foundational models utilizing Amazon SageMaker. Amazon SageMaker could be a completely overseen benefit given by AWS, particularly planned for building, preparing, and conveying machine learning (ML) models. By creating foundational models custom-made to healthcare scenarios, GE Healthcare points to quicken the advancement and arrangement of web-based therapeutic imaging applications and progress proficiency, interoperability, and client involvement by coordination these foundational models. GE Healthcare's generative AI applications will be profoundly coordinates with Amazon HealthLake and Amazon HealthImaging, empowering clients to rapidly and safely analyze different sorts of understanding information, subsequently moving forward clinical effectiveness and improving understanding care quality.
Globally, the healthcare sector generates nearly 30% of the world's data, yet as much as 97% of this data remains underutilized due to its unstructured nature. The collaboration between AWS and GE Healthcare will focus on developing multimodal foundational models to deeply analyze vast amounts of unstructured healthcare data, including medical records, reports, and imaging, providing precise and highly adaptable analytical results for diverse healthcare applications. Compared to traditional machine learning models, these new models can simultaneously process and interpret data related to various diseases and medical tasks, offering more flexible support for clinical decision-making.
GE Healthcare has long been committed to AI research and development. In the U.S. Food and Drug Administration (FDA) AI device authorization listings, GE Healthcare has ranked first for three consecutive years, with a total of 72 authorizations. One of GE Healthcare's recent research achievements in the field of foundational models is a project focused on advanced ultrasound image segmentation tools. Utilizing foundational model technology, this tool has demonstrated exceptional accuracy in separating and identifying anatomical structures, with a precision rate exceeding 90%, requiring minimal human intervention. A significant advantage of this tool lies in its outstanding generalization capability, enabling precise segmentation of various anatomical structures, including fetal heads and lesion areas in breasts, even though these structures were not included in the original training data of its foundational model. This breakthrough capability showcases the model's vast potential for applications far beyond its initial training scope.
Phone