As artificial intelligence continues to evolve the medical field, UofL is investigating how AI could help improve patient outcomes during heart surgery.
A $750,000 grant from the American Heart Association will allow researchers to advance AI specifically for acute kidney injury and complications during or following cardiac surgery.
Acute kidney injury can result in increased mortality or persistent kidney dysfunction and, because it has a wide variety of contributing factors from patient-specific conditions to procedure complexity, this issue can be difficult for physicians to predict and prevent.
The project is a joint effort between UofL researchers from the School of Medicine, School of Public Health and Information Sciences, the J.B. Speed School of Engineering, UofL Health and researchers at SUNY Buffalo, Georgia Institute of Technology and Baylor Scott & White Heart and Vascular Institute.
The team will innovate machine-learning AI models to analyze detailed, clinical patient data and develop a personalized risk prediction and decision-making process for managing kidney injury in heart surgery patients. They then will validate the process using independent databases and clinical trials at UofL Health.
UofL’s Jiapeng Huang, professor and vice chair of the anesthesiology and perioperative medicine department, is principal investigator for the project. As a cardiac anesthesiologist at UofL Health, he also sees numerous patients who deal with acute kidney injury.
“Our goal is to use AI and machine learning methodology to do two things. One, to predict in real time when the patient might develop acute kidney injury or if the patient will be at risk for acute kidney injury,” he said. “The second thing is to develop a clinical decision-support system to help the clinicians do the right thing for the patients at the right time to reduce chance of acute kidney injury after heart surgery.”
While Huang and UofL faculty member Bert Little focus on the clinical procedures and decision-making process, Lihui Bai, professor of industrial engineering at the Speed School, Xiaoyu Chen, assistant professor of industrial and systems engineering at SUNY Buffalo and George (Guanghui) Lan, professor of industrial and systems engineering at Georgia Institute of Technology, will work with a team of engineers to build the AI technology. The tech will allow physicians to use patients’ clinical information before, during and after surgery to inform physicians of the best sequence of treatment for patients to reduce the chance of kidney injury after heart surgery.
For the last 10 years, AI has been used in the medical field to analyze large health care data. AI can more easily recognize patterns than the human eye or brain, according to Huang, and can be a significant benefit to patient outcomes.
“This is one of those research (projects) that will benefit patients directly,” he said “Acute kidney injury happens in about 25% of patients after cardiac surgery. This study aims to protect patients from acute kidney injury after heart surgery.”
The three-year project, which is currently in phase 1, began in July 2023. During this early phase, the team is establishing the database and prediction model. In year three, clinical trials conducted at UofL Health will be used to determine whether the predictive modeling and clinical decision support system will reduce the rate of acute kidney injury after cardiac surgery.
UofL Health is an excellent partner for this project as it is one of the premier cardiac programs in the nation, according to Huang. It was responsible for the first heart transplant in the state of Kentucky, as well as many innovations in artificial heart pumps. UofL Health cardiovascular surgeon Siddharth Pahwa and cardiologist Dinesh Kalra, for example, are involved in other studies, including cardiac imaging and data collection in addition to patient care.
“UofL Health always focuses on improving patient safety and outcomes,” Huang said. “UofL faculty and researchers are perfect partners to perform clinical studies to advance our knowledge and benefit our patients at UofL Health.”
