A new research project funded by the National Science Foundation aims to develop implantable medical devices that can learn from a patient's body and adapt automatically, potentially reducing the need for frequent follow-up visits and improving care for a wide range of conditions.

According to Syracuse University, the $1 million grant supports a multi-university effort led by Alex K. Jones, chair of the Department of Electrical Engineering and Computer Science in the College of Engineering and Computer Science (ECS) and the Klaus Schroder endowed professor of engineering at Syracuse. Jones is a principal investigator on the project, which also includes researchers from the University of Notre Dame and the University of Texas at San Antonio.

The project targets devices such as pacemakers, implantable cardioverter defibrillators, deep brain stimulators, cochlear implants and sleep apnea implants. Currently, these devices are typically programmed at the time of implantation and can often only be adjusted during follow-up visits with a doctor, even though a patient's condition can change from day to day or evolve over years as they age.

"Implantable medical devices must operate reliably for years while responding to changes in a patient's physiology," Jones said in a statement. "Our challenge is to enable increasingly sophisticated intelligence and adaptation within extremely constrained energy budgets, extending device lifetimes while improving patient outcomes and reducing the need for invasive replacement procedures."

A central challenge, the researchers explained, is developing lightweight artificial intelligence models that can interpret physiological signals, recognize when conditions are changing, and adapt safely under uncertainty. The team plans to combine these algorithms with energy-efficient computing architectures, including in-memory computing, to minimize the energy required for continuous monitoring and learning. By co-designing the algorithms and hardware, the researchers aim to create implantable systems capable of evolving their behavior over time without compromising safety or device lifetime.

Jones brings expertise in energy-efficient computing architectures, in-memory computing and hardware-software co-design to the project. Yiyu Shi, professor of computer science and engineering at Notre Dame, is the lead principal investigator. Jinjun Xiong, dean of the College of AI, Cyber and Computing at the University of Texas at San Antonio, is also a collaborator.

For students, the research could have several implications. The project includes an education and outreach component that will produce open-source educational materials, support undergraduate and graduate training, and offer research experiences in artificial intelligence, embedded systems and semiconductor technologies. Outreach will also include K-12 students and military veterans, and the team plans to release public datasets and benchmarking platforms.

Julie Hasenwinkel, interim dean of the ECS, said the work demonstrates what is possible when artificial intelligence is integrated with medical device design. "Devices that can learn from and adapt to an individual patient have the potential to improve care in ways that will continue to evolve for many years to come," she said.

The researchers also noted that self-evolving medical devices could help reduce health disparities by adapting to people whose individual characteristics differ from those represented in conventional device designs. Greater autonomy could also reduce the need for clinical follow-up appointments, which may be particularly beneficial for patients in remote or underserved areas.

The award was made through the NSF's Computer Systems Research program in the Directorate for Computer and Information Science and Engineering.