Xiang Zhong, Ph.D., Uses AI to Transform Intensive Care Delivery

Xiang Zhong, Ph.D., assistant professor of Industrial & Systems Engineering (ISE) at the University of Florida recently received funding from the National Science Foundation in support of her research in developing a causal AI digital twin framework to tackle the challenges of critical care delivery. 

According to Dr. Zhong and her team, there have been long-standing deficiencies in critical care practices in hospitals on a global scale. Due to the COVID-19 pandemic, these conditions have only regressed. As a result, new methods and strategies are needed to facilitate timely and accurate interventions during intensive care procedures. 

Dr. Zhong’s research is a collaborative project with the Mayo Clinic and aims to re-envision intensive care delivery processes through three tasks: (1) learning a causal AI model to underpin the clinical pathway of critically ill patients; (2) investigating optimal treatment of these patients during the first 24 hours (“golden hours” for treatment and intervention): (3) and enabling systems-level interventions through a digital twin framework.  

As an initial case study, the team has chosen patients suffering from sepsis, an extreme reaction of the body from an infection, to test the developed prototype. The researchers will utilize computationally efficient algorithms to build what’s known as a dynamic Bayesian network, or a probabilistic graphical model. The model will learn treatment rules based on electronic health record data for integration with structures informed by expert rules thereby harnessing both human and artificial intelligence. This customized model architecture will process information to support multiple simulations in a flexible and efficient manner.  

The goal of this research is to contribute to the advancement of national health and welfare and support critical care delivery and facilitate the realization of “Healthcare 4.0.”Healthcare 4.0 is considered the next digital revolution in healthcare, with a focus on gathering data to help practitioners make more informed healthcare decisions that are more cost-efficient. 

“Through this research, a diverse group of students and clinical fellows will receive a blend of interdisciplinary training in machine learning, systems engineering, and critical care medicine. This causal AI digital twin framework will be a critical leap forward in support of a more efficient medical education, and eventually less error-prone bedside decision making,” said Dr. Zhong.