Research Team Taps AI for Better Cardio Care
Technique Could Allow Doctors, Patients to Match ‘Gold Standard’ Monitoring at Home
A UMD-led research team is using artificial intelligence in a project to boost the reliability of data from a low-cost, user friendly method to monitor for heart disease home to equal that of the less-accessible clinical gold standard.
Illustration by iStock
The nation’s leading cause of death, cardiovascular disease, is often referred to as the “silent killer” because detecting symptoms early is so difficult. Current testing methods are time-consuming, expensive and inaccessible, especially for low-income and other disadvantaged patients. But what if an alternative were literally at your fingertip?
A University of Maryland expert in digital forensics and signal processing has received a $1.2 million award from the National Science Foundation (NSF) to use artificial intelligence (AI) to develop a heart monitoring method that’s as reliable as the gold standard used in hospitals and clinics—an electrocardiogram (ECG)—but as convenient as a device that can be worn at home.
Min Wu, a professor in electrical and computer engineering with a joint appointment in the University of Maryland Institute for Advanced Computer Studies, is leading the four-year project. She is collaborating with Sushant Ranadive, an expert in cardiovascular physiology and assistant professor of kinesiology in UMD’s School of Public Health.
The team is developing an innovative way to understand the relationship between results from ECG—where electrodes are placed on the patient’s chest—and those from a method known as photoplethysmogram (PPG), which measures cardiac activity by monitoring changes in blood volume beneath the skin through a sensor that could be worn on a finger. (While it’s currently possible to obtain instant ECG data through a smartwatch or special smartphone attachment, these methods are impractical for long-term monitoring, the researchers said.)
However, while PPG is cheaper, more convenient and more accessible than traditional ECG testing, it provides less direct information on cardiac activity and is not as well understood by researchers or clinicians. The main goal of the NSF-funded project is to compensate for that gap by using AI to reconstruct ECG-quality results with PPG data, said Wu.
Her team plans to work closely with Clifton Watt, M.D., a cardiologist at of the University of California, San Francisco, to transfer the substantial ECG medical knowledge base to the PPG domain.
Using an existing dataset from other researchers, her team has already carried out preliminary studies on a few hundred hospitalized patients. The NSF funding will allow exploration of a wide range of research questions the team hopes will result in a user-friendly self-monitoring system.
“The bridge between ECG and PPG enabled by explainable AI could bring unprecedented opportunities to expand smart health knowledge and benefit public health,” Wu said.
Topics: