The collaboration will develop AI models and applications for healthcare professionals, with the goal of improving healthcare for us all.
Nearly one-third of digital information is generated in healthcare, yet 97% of it is unavailable to doctors and clinicians because it not only takes a variety of unstructured formats, from handwritten notes to images, but also tends to be locked in data silos across the world.
Today, AWS and GE HealthCare announced a collaboration that aims to unlock critical information and unlock possibilities for a new era of health using industry-specific artificial intelligence (AI) foundational models (FMs) and new applications. GE HealthCare announced plans to train and deploy new clinical FMs on AWS machine learning and generative AI technology with the goal of empowering healthcare providers not only to improve existing protocols and workflows, but also to invent entirely new approaches to delivering better care to all patients.
As with all things AI, collecting, analyzing, and leveraging the right data is key to unlocking new insights and solutions. In healthcare, relevant data includes things like doctor's notes, x-rays, and any other test results or observations collected during a consultation.
“Healthcare systems collect vast amounts of data points about patients over their lifetime. Unfortunately, only a small amount of that data can be efficiently and securely accessed to inform patient diagnosis, prognosis, and treatment.“GE Healthcare's Global Chief Scientific and Technology Officer, Dr. Taha Kashout, said:Until recently, the technology to securely and efficiently aggregate, analyze, and interpret this data did not exist..”
Learn about AWS and its growing range of powerful, secure AI capabilities and services.
GE HealthCare will use Amazon Bedrock to leverage world-leading foundational models, train new models and develop AI-powered applications to help hospitals and clinics access and gain insights to improve patient care based on a holistic view of data rather than a partial view. To support the development of these future AI applications, GE HealthCare's in-house developers will use Amazon Q Developer and Amazon Q Business to explore the intersection of multi-modal clinical and operational data with the goal of reducing physician burden, enabling more personalized care, and increasing efficiencies.
“GE Healthcare is using AWS to plan its cloud deployment to deliver more personalized, intelligent and efficient care.“AWS CEO Matt Gurman said:GE Healthcare is putting generative AI at the center of its innovation, accelerated by investments in healthcare-specific cloud services and generative AI capabilities that offer best-in-class security, data privacy and access to the latest cutting-edge foundational models..”
Towards more personalized care
The foundational model GE Healthcare creates can be adapted and tweaked to specific medical tasks, says Kass Hout, a trained interventional cardiologist who has held a variety of technology and medical positions at Amazon, the FDA and the CDC. That makes the model an ideal foundation for creating innovative applications, he says. And because the model is on the cloud, doctors can access it online from anywhere, whether in their office or remotely.
For example, we could build tools that help doctors quickly review a patient's entire medical history to diagnose the cause of recent symptoms. Cloud-based generative AI applications built using these foundational models can collate and flag notes, recordings, and images for doctors in minutes or seconds instead of days for a human, allowing doctors to make accurate, personalized diagnoses. Similarly, these apps can provide a more comprehensive and detailed view of a patient's medical history, helping clinicians detect emerging issues earlier and flag precision treatments.
Kasshout believes that these foundational models and the resulting applications could lead to “significant” breakthroughs in the broader diagnosis of medical conditions and the provision of specialized treatments. For example, GE Healthcare recently fine-tuned a research foundational model used in ultrasound images to identify anatomical structures without direct training, highlighting how generative AI can help accelerate innovation in healthcare. If such technology is eventually introduced into clinical practice, it could allow cardiologists to easily examine the structure of a diseased heart in a given patient, for example. It could also shorten the development cycle for clinical applications from years to months.
In short, this technology has the potential to bring about new standards of care that are more predictive and preventative, rather than reactive.
Protecting Patient Privacy with AWS Security
Kass-Hout said AWS' cloud infrastructure and advanced machine learning (ML) services also provide the processing power and security needed to manage the large volumes of complex data that healthcare professionals process every day, meaning GE HealthCare can develop and offer to customers generative AI tools that comply with privacy regulations such as the Health Insurance Portability and Accountability Act (HIPAA).
Kass Hout, who has spent decades in the healthcare industry, says the ability to collect, aggregate, analyze, and leverage all of a healthcare organization's data was a pipe dream for most of his career, but he believes recent advances in the models underlying generative AI are making that dream a reality.
“We haven't seen such progress since the advent of the Internet.,” He said. “These foundational models have the potential to revolutionize healthcare data and make precision healthcare analytics as ubiquitous as the web.“
Learn more about GE HealthCare’s digital solutions and healthcare transformation here.
To learn more about AWS generative AI capabilities, please visit here.
