Accelerating healing and improving health outcomes for 467,000 service member and civilian patients with AI-powered precision medicine
$10 billion in potential annual civilian cost savings
Weeks instead of years for analyses fueling biomarker discovery and biological insights
57% reduction in wound closure complication rates using WounDx tool on Google Cloud
Researchers and clinicians at Uniformed Services University leverage a comprehensive suite of Google Cloud services—including BigQuery, Cloud SQL, Compute Engine, and VertexAI—to speed discoveries that accelerate patient recovery, yielding $10 billion in potential cost savings annually.
Researchers and clinicians at Uniformed Services University leverage a comprehensive suite of Google Cloud services—including BigQuery, Cloud SQL, Compute Engine, and VertexAI—to speed discoveries that accelerate patient recovery, yielding $10 billion in potential cost savings annually.
Increasing speed and collaboration for medical research in the cloud
When people decide to dedicate themselves to the U.S. military, they deserve all of the support that the country can provide, including top medical care. The U.S. government officially established the Uniformed Services University of the Health Sciences (USU) in 1972 with the specific mission of educating and training medical professionals to treat U.S. military members and their families around the world. The university provides a curriculum that blends medical education and clinical practice with innovative research to create the next generation of healthcare.
Many of USU’s research programs focus on issues that impact military personnel, such as critical injury care. Dr. Eric Elster, School of Medicine Dean and EVP for Medical Affairs at USU, founded the Surgical Critical Care Initiative (SC2i), which leverages Google Cloud technology to conduct research and develop critical care tools and strategies based on precision medicine.
Data is the driving force behind SC2i research. Researchers and clinicians collaborate to gather and analyze large volumes of biological and medical data to find patterns that can lead to personalized treatments. Then SC2i turns these insights into clinical decision support tools (CDSTs) that help physicians quickly identify the right treatment at the right time for every patient.
After years of working on individual workstations, the program decided to adopt Google Cloud solutions to accelerate collaboration, research, and development of CDSTs.
We have harnessed Google Cloud BigQuery, Vertex AI, and Gemini models to build upon our legacy of innovation at USU. By establishing a single vision for how we approach compute and storage, and with AI integrated across our platforms, we are accelerating research, improving care, and returning time to clinicians.
Sean R. Baker
CTO and SISO, Uniformed Services University
With Google Cloud, SC2i can provide its teams with a strong data foundation alongside powerful AI and machine learning models to develop clinical care tools with the potential to impact 467,000 patients and save civilians $10 billion every year in the United States.
Training machine learning models for better patient outcomes
With Google Cloud, SC2i breaks down silos between researchers and encourages collaboration for faster development and performance of clinical decision support tools. Previously, staff worked with flat spreadsheets and databases available to only one user at a time. Now multiple researchers can work simultaneously to analyze over 100 million data elements using BigQuery.
This solid data foundation for collaboration and AI supports their collaborative approach, increasing efficiency and speed of discovery, potentially from years to months. SC2i has dramatically cut cycle times from years to weeks for transcriptomics and other proteomic analyses that fuel biomarker discovery and biological insights.
WounDx will cut wound closure complication rates by more than half, saving approximately $60,000 per patient.
Dr. Seth Schobel
Scientific Director of SC2i, Uniformed Services University
With Vertex AI, researchers can connect data to the most powerful AI and machine learning models available, including Gemini. This accelerates research by helping scientists analyze the data to identify key clinical characteristics and biomarkers faster. Analysis that used to take weeks or months can be repeated by researchers in the cloud in just weeks or days.
“AI tools are becoming embedded into everything we do,” says Dr. Elster. “They’re proving to be game changers, helping us accelerate innovation and improve efficiency in medicine.”
Developers use data insights to create diagnostic tools for clinicians in Google Cloud. BigQuery helps to ingest and analyze data faster, improving performance for near real-time results. Working in Google Cloud, developers can also validate the environments to meet medical device standards, such as ISO 13485, to make sure that the technology is ready for clinical use.
One of the first decision support tools from SC2i to go to clinical trials is WounDx, which helps surgeons optimize the timing for wound closure. When working on a patient with a critical traumatic injury, clinicians can collect a sample from the wound and enter the diagnostic results into a secure digital interface. WounDx applies its machine learning model to predict whether the wound is ready for closure, or if it should be left open longer to promote healing. A Gemini model then generates a report that summarizes the recommendations for clinicians.
SC2i expects that using patient data to determine wound closure timing, rather than relying on standard timings, will reduce the chances of wound dehiscence—a complication where a surgical wound reopens. Research suggests that WounDx will reduce wound dehiscence rates from 23% to just 10%, for substantial cost savings of around $60,000 per patient. More importantly, it has a big impact on patient recovery with decreased pain, reduced complications, and fewer hospital visits.
Cloud computing and AI: the future of medicine
SC2i researchers have around 10 additional clinical decision support tools in various stages of development. An upcoming CDST will use similar workflows as WounDx to provide early detection for three of the biggest concerns during recovery: pneumonia, blood clots, and acute kidney injury. By identifying issues earlier, SC2i aims to reduce complications and risks for patients.
While USU uses Gemini to support research and development, many staffers also use Google’s generative AI capabilities to support administration and paperwork, such as drafting emails and reports. This gives time back to scientists and educators to focus on what’s really important: developing the future of medical care.
The medical field is poised for dramatic change, and a lot of that will come from cloud computing and AI. With technology supporting physicians, we can reduce the time people spend in hospitals, see more patients, and improve quality of care to help get patients back to their lives faster.
Dr. Eric Elster
School of Medicine Dean and EVP for Medical Affairs, Uniformed Services University
Uniformed Services University of the Health Sciences (USU) is the professional school of medicine and health science for the U.S. federal government and military services. Since its founding in 1972, it has produced more than 10,000 healthcare provider professionals, scientists, policymakers, and educators. The university’s research has resulted in 744 life-saving patents.
Industries: Government and Public Sector, Healthcare
Location: United States