AI-Powered Tools Aim to Transform Chronic Pain Management for Rural Older Adults
核心洞察
Virginia Tech researchers received a $460,260 National Institute on Aging grant to use artificial intelligence to improve pain assessment and clinical decision-making for rural older adults with chronic pain (搜索).
An estimated 36 percent of U.S. adults over age 65 experience pain most days or every day, with rural communities facing reduced access to pain specialists and healthcare resources.
The interdisciplinary team will apply machine learning and natural language processing to extract pain information buried in clinical notes and build risk dashboards for complications such as depression (搜索) and cognitive impairment (搜索).
Older adults commonly struggle with chronic pain (搜索), with an estimated 36 percent of U.S. adults over the age of 65 reporting pain most days or every day, according to the Centers for Disease Control and Prevention. That burden can be even more challenging for those living in rural communities, where access to pain specialists and healthcare resources is more limited. Huaiyang Zhong, assistant professor in the Grado Department of Industrial and Systems Engineering at Virginia Tech, is working to address this gap using artificial intelligence methods to improve pain assessments, identify risks, and support better clinical decision-making for older patients in rural communities.
The effort is supported by a $460,260 grant from the National Institute on Aging at the National Institutes of Health. "I became interested in chronic pain (搜索) because it's incredibly common, but also incredibly complicated," Zhong said. "Pain is not just a single diagnosis, and it's not a number on a scale like a lot of medical diagnoses. Pain itself can affect mobility, mental health, sleep, cognitive functions, and overall quality of life. It's a multidimensional thing."
Extracting Pain Data Hidden in Clinical Notes
The primary goal of the research is to understand pain better. Zhong hopes to learn how chronic pain (搜索) evolves, which patients are increasingly vulnerable to poor outcomes, and how healthcare systems can intervene effectively. That work starts with data. Valuable information about a patient's pain experiences may be buried in clinical notes, as opposed to structured electronic health records. The researchers will use machine learning and natural language processing to extract that information systematically, then use the data to create risk dashboards that allow clinicians to recognize when a patient has a greater chance of complications, such as depression (搜索) or cognitive impairment (搜索).
"I ultimately want to help clinicians move toward more personalized pain management," Zhong said. "This means understanding not just how much pain somebody has, but the broader health context surrounding that pain."
An Interdisciplinary, Systems-Level Approach
To ensure the research on his first National Institutes of Health grant has the most clinical impact, Zhong has assembled an interdisciplinary team of faculty and clinicians from medicine, computer science, statistics, and whole health. Collaborators include Robert McNamara, associate professor in the Virginia Tech Carilion School of Medicine (VTCSOM) Department of Psychiatry and Behavioral Medicine and clinical psychologist at Carilion Clinic; Tina Savla, director of the Whole Health Consortium; Robert Trestman, chair of psychiatry and behavioral medicine at VTCSOM and Carilion Clinic; Anita Kablinger, program director and professor of psychiatry and behavioral medicine at VTCSOM; Margaret Rukstalis, associate professor of psychiatry and behavioral medicine at VTCSOM; Alexandra Hanlon, director of the Center for Biostatistics and Health Data Science; Elizabeth Russo-Stringer, assistant professor in surgery at VTCSOM and chief of interventional pain management in Carilion Clinic's Department of Surgery; Kwok Tsui, professor at the University of Texas-Arlington; Donald Penzien, professor at Wake Forest University School of Medicine; and Xuan Wang, assistant professor in Virginia Tech's Department of Computer Science.
"Our team is grateful to be awarded support to investigate the medical complexities experienced by rural older adults experiencing chronic pain (搜索)," said McNamara. "We foresee this work leading to early, actionable insight for providers in rural areas, enabling appropriate intervention and referral, and ultimately improving quality of life for this vulnerable population."
Zhong believes industrial and systems engineering (ISE) is uniquely positioned to help solve overarching healthcare problems, given that the discipline is fundamentally about improving complex systems. "Machine learning can tell us which patients are at elevated risk, but as ISE researchers, we ask the next questions: What should we do with that information? How should limited healthcare resources be allocated? How does this information feed into clinical workflows? How does using this actually improve clinical outcomes?" Zhong said. "That decision-oriented, systems-level perspective is what ISE uniquely brings to healthcare."
Digital Tools as Complementary Care
The Virginia Tech work aligns with a broader movement toward digital and virtual tools for chronic pain (搜索) management. Eric Anderson, MD, neurologist and chief medical officer at Lin Health (搜索), has developed care models in the digital and virtual space that address the biological, psychological, and social drivers of chronic pain, focusing specifically on primary pain or neuroplastic pain. "All pain signaling comes from the brain," Anderson said. "When we touch something, mechanical receptors tell us there's pressure, temperature, something noxious. Our brain processes that and determines what to do with it. In primary pain, this becomes somewhat maladaptive. The brain becomes overprotective, registering things that are no longer pain or that could be pain as pain."
Anderson emphasized that digital tools should complement, not replace, existing medical care. "Digital tools and virtual programs should be bolted on to a service that extends the reach of an existing medical program," he said. "Augment, not replace."
In the chronic pain (搜索) treatment space, neuromodulation devices can "rewire pain signaling." These therapies, which include spinal cord and deep brain stimulation, work via wireless implants that can respond in real time to reduce pain. Such approaches could be further coupled with adaptive AI that could learn patient patterns and help provide relief from pain and even prevent flare-ups before they manifest. In 2022, the US Food and Drug Administration approved virtual reality systems as a chronic pain management tool, and a crossover study published in April 2025 showed that telehealth virtual reality interventions can reduce chronic pain.
Dimitri Souza, MD, a physician at Western Reserve Hospital Center for Pain Medicine, described watching a patient with sickle cell disease (搜索), considered among the most painful conditions, using a virtual reality system. "He was in a different world, flying between the hills. He was so excited and happy and there was no sense of pain on his face. It's a really powerful tool. And it is available. Like the endorphins and pain protection systems we all have. They're there [for us to use]. We just need to learn how."
As AI and other digital advances become more widespread, a future that eases the effects of chronic pain (搜索) becomes possible and feasible, particularly for underserved populations such as rural older adults who stand to benefit most from tools that extend the reach of specialized care.
