A team of researchers at The University of Texas at Arlington has received a grant from the Cancer Prevention and Research Institute of Texas, known as CPRIT, to tackle the critical issue of low physical activity among cancer survivors. Regular exercise can significantly improve survivors' quality of life and reduce their risk of death, yet up to 84% don't get enough to see these benefits.
UT Arlington researchers are developing a personalized approach to encourage and support physical activity among cancer survivors. Their smart system will use AI and machine learning to send messages to survivors tailored to them in real time using data from their wearable devices and smartphones. The goal is to deliver the right message at the right moment for maximum impact.
The research team includes Yue Liao, assistant professor in UTA's public health program in the Department of Kinesiology; Grace Brannon, associate professor in the Department of Communication; Chengkai Li, professor and associate chair in the Department of Computer Science and Engineering; and Maria Chang, a collaborator at MD Anderson Cancer Center in Houston.
The two-year project will develop an algorithm that can create timely, personalized messages then test them on 15 participants.
We can't always predict what people will be doing, and we only have a certain amount of information to develop intervention messages for patients. If we start to think about all the different scenarios recovering cancer patients might be living day to day, that'll involve developing thousands and thousands of personalized messages, which a human just can't do. As the team envisioned how to work past this, AI was becoming a more popular tool, and we thought it could really help us scale this research."
Dr. Yue Liao, assistant professor in UTA's public health program in the Department of Kinesiology
This project expands on current research by Liao and Dr. Brannon. The pair connected with Dr. Li through UTA's Center for Innovation in Health Informatics to tap his expertise in AI.
"AI is now the driving force of almost every scientific field-it's a pillar of our society's advancement," Li said. "I'm excited to be part of this project because it is truly interdisciplinary work and I'm keen on collaborating with experts outside of my field. Bringing together different perspectives and working together at the intersection of multiple fields can make great breakthroughs in our different areas of study."
With little research currently focused on this area, the UTA team's work could help transform how behavioral interventions are designed and delivered. By recognizing patterns in user data, AI can select the most appropriate message for each individual.
"In our initial pre-pilot testing, we had about 30 participants, and using humans to personalize the intervention messages was manageable," Brannon said. "However, if we're trying to get personalized intervention messages to thousands more people, we need something that is more cost- and time-effective. AI allows for more adaptations with its algorithms, and it helps this project become much more scalable and cost-effective while being personalized."
This kind of interdisciplinary collaboration enables the team to explore solutions beyond the scope of their individual fields-unlocking innovations with broader impact.
"There were so many factors that I needed to consider in my research, but I just didn't have all the expertise," Liao said. "At UTA, I've been able to pull experts from different fields and not feel limited. Being part of an interdisciplinary team has definitely helped broaden my research scope."
Researchers identify cellular hotspots that trigger ferroptosis