Advocate Aurora Research Institute uses AI and machine learning to transform surgery

A research project with KelaHealth harnesses predictive analytics to evaluate robotic and other surgical techniques with an aim to reduce variability and improve patient outcomes

Share this story:
Advocate Aurora Research Institute uses AI and machine learning to transform surgery

Advocate Aurora Research Institute is collaborating with health care predictive analytics company KelaHealth, leveraging its advanced machine learning models and predictive insights to transform surgical care at midwestern Aurora Health Care and Advocate Health Care sites.

The research project will examine how much variation exists in surgical and patient outcomes to better understand ways to personalize surgical care for each patient.

“Until recently, surgeons throughout history have relied on their education, experience and intuition to make care decisions and select surgical techniques,” said Debra O'Connor, DO, Vice President of Quality and Clinical Effectiveness for Advocate and Aurora and co-principal investigator for the research project. “But now, our surgeons have an extra advantage at their disposal – vast amounts of data from the electronic health record and other surgical data sources that can be used to reduce any unwelcome variation between surgical approaches, departments, hospitals and even individual surgeons.”

The researchers will feed data from the health system’s electronic health record and new sources, such as robotics data, into machine-learning software from KelaHealth, the study’s sponsor. KelaHealth’s proprietary Surgical Intelligence Platform uses AI to create risk-prediction models using historical data from surgeries. Researchers can then analyze the patient-specific predictive insights to better determine which surgical methods would have resulted in the best outcomes for each of those past patients.

“Our goal is to improve patient outcomes by analyzing and understanding the varied practices by surgeons coupled with the unique needs of the patient,” said Gary Chmielewski, MD, thoracic surgeon and Robotics Committee Co-Chair for Advocate and Aurora. “It’s possible that, say, gallbladder surgery is best performed robotically for all of our patients or patients with specific health risks. This project will give us these kinds of insights, which can help us guide surgical decision-making, not just as a learning health system, but right down to the tendencies of individual surgeons, in the context of understanding how to deliver better care for patients.”

Once the Surgical Intelligence Platform has used this data to create a baseline, the researchers will continue to use the prediction model over time to review changes in surgical approaches and performance. The research project has the potential to advance health equity by personalizing care to each patient’s need, reducing unwarranted variation and accelerating medical breakthroughs.

A key element of the project is the inclusion of robotic data from the health system’s more than 37 Intuitive da Vinci Systems, which are currently used by 284 Advocate and Aurora surgeons in Illinois and Wisconsin.

“Computer-assisted surgery has been digital since its inception, and it’s this kind of data-driven research that can help improve patient outcomes,” said Brian Miller, Executive Vice President and Chief Digital Officer for Intuitive. “In many ways we’re just scratching the surface of how data from all sources – enabled by AI and machine learning – can surface to health care providers actionable insights for each patient’s unique risk factors, thereby helping health care providers plan for the best surgical results.”

“With surgery comprising 50% of Medicare spending, it is imperative for health systems to explore new ways to improve their stewardship of these resources. The modern surgical process generates tremendous amounts of data – from robots, medical devices, instruments and sensors – yet its potential remains untapped,” said Bora Chang, MD, CEO of KelaHealth. “Our core competence is translating these data points into predictive insights and effective interventions enabling hospitals to operate smarter, reducing the cost and improving the quality of surgical care.”

To learn more about our research, visit aah.org/research.

View More News Releases

About KelaHealth

KelaHealth is a surgical data analytics company that delivers patient-specific predictive insight and risk stratification software – The KelaHealth Surgical Intelligence Platform – to help improve care quality and clinical operations. Founded by a group of surgeon-scientists, KelaHealth’s goal is to harness AI and machine learning algorithms to reduce patient surgical complications and improve outcomes. For more information about KelaHealth, visit www.KelaHealth.com.

About Advocate Aurora Research Institute

Advocate Aurora Research Institute is a not-for-profit, limited liability company of Advocate Aurora Health. Advocate Aurora has emerged as a national destination for patient-centered bench, translational and clinical research, and the Research Institute unifies the innovative research efforts throughout the health system. Advocate Aurora researchers focus on rapidly translating new discoveries from the scientist’s bench to the patient’s bedside and into the community we serve to improve options and outcomes that change not only the lives of individuals, but transform the health of populations.