Reimagining health monitoring for everyone

By Celine Gomes

Across the United States, 12% of the population lives with at least one chronic condition, which can place a heavy burden on families, hospitals, and the healthcare system at large. What if a single device could help reduce hospital admissions, free up critical resources, and lower health costs, while working equitably across all skin tones?

Associate Professor Valencia Koomson of the Department of Electrical and Computer Engineering is leading that charge. Alongside PhD candidates Blessing Kolawole (Computer Science) and Ravi Durbha (Electrical and Computer Engineering), Koomson’s team has developed an AI-powered health monitoring device designed with inclusivity at its core. 

For over 20 years, Koomson has dedicated her lab to advancing research at the intersection of engineering and medicine. Her work centers on developing non-invasive biomedical imaging and diagnostic technologies that deliver clinical insight without the need for surgery. Her latest project? A lightweight, wrist-worn device that monitors key health indicators—blood pressure, heart rate, blood oxygen levels, and more—in real time.

Designing an AI-powered pulse oximeter 

“The goal is to develop a wearable medical device capable of measuring a multitude of parameters. Most medical devices in current circulation can only measure one or two,” Koomson explains. From that goal stemmed critical questions: “How do we build the hardware to capture physiological data non-invasively, and once we do, how do we analyze it? What can we extract from it?” The answers lie between advanced algorithms and precise hardware engineering.

This is where Kolawole and Durbha come in. The project initially presented a range of technical challenges, particularly in how light interacts with melanin. “Optical signals are absorbed and scattered differently depending on skin pigmentation,” explains Koomson. That means devices relying on light-based sensors often produce inconsistent results across different skin tones. To overcome these limitations, Durbha combined two complementary technologies. “One aspect of the innovation focused on enhancing pulse oximeter performance through advanced signal processing algorithms and refined calibration models. The other involved integrating a color sensor to actively account for skin tone variations, enabling more equitable and accurate oxygen saturation measurements across diverse populations,” he explains. Durbha led the development of the signal processing pipeline, designing algorithms that continuously assess the device’s accuracy and reliability in real time. 

Kolawole’s work centered on AI integration for on-device inference, particularly to estimate vital signs such as blood pressure and arterial stiffness using physiological data gathered from sensors. Kolawole examined existing and incoming data for potential biases—ensuring the algorithms were trained on diverse, representative data sets to avoid systemic inaccuracies. “We had to identify not just where bias currently exists, but where it could be introduced in future models,” she explains. 

Together, their interdisciplinary collaboration brought hardware and AI into harmony. The result: a cost-effective, inclusive, and accurate device engineered to pave the way toward more equitable healthcare outcomes. With it, people could detect symptoms early, get answers quickly, and avoid unnecessary hospital visits that often involve hours of waiting and high out-of-pocket costs.

Award recognizes project’s potential 

Their work was recently honored with the Stephen and Geraldine Ricci Interdisciplinary Prize, awarded for the project’s potential to transform a research discovery into a real-world application that benefits society. Presented in early April, the honor demonstrates the team’s entrepreneurial spirit and interdisciplinary engineering excellence. 

Koomson, who also advised the 2014 Ricci-winning team, emphasized the importance of the prize: “It’s a great boost for students who work so hard on these projects—especially when they’re trying to collaborate across disciplines, which adds another layer of complexity.” 

For Kolawole, the recognition reinforced the broader purpose behind their work: “The award is a validation of the work we’re doing and the gravity of the impact we hope this device will make in the world.” Durbha added, “This work pushed me into a completely new domain: medical technology. But even with different backgrounds, with some effort, you can come together and build something truly innovative that can help society in a meaningful way.”

Technology continues to transform how we communicate, travel, connect, and care for ourselves and each other. The future of biotech lies in making personalized, preventive healthcare a reality for everyone, everywhere. Koomson, Kolawole, and Durbha’s work is a powerful reminder that when we challenge assumptions and collaborate across disciplines, we don’t just advance science, we reimagine what’s possible.