AI’s application to the study of the human body has yielded another potential development, one that researchers say could eventually give neurologists a powerful new tool in the fight against Alzheimer’s disease, stroke and other neurological disorders.
Beyond allowing the body to rest, deep sleep serves another critical function: It helps the brain clear metabolic waste through what scientists call the glymphatic system. During that process, water-like fluids known as cerebrospinal fluid and interstitial fluid circulate through the brain, helping wash away proteins and other byproducts, including amyloid beta, which has been linked to Alzheimer’s disease.
Understanding exactly how those fluids move through the living brain has long posed a challenge for researchers. Directly measuring fluid circulation deep inside the brain would require invasive procedures.
AI is helping overcome that obstacle. Douglas Kelley, a professor of mechanical engineering at the University of Rochester, and colleagues from the University of Rochester, Brown University and the University of Copenhagen have developed a new approach that combines MRI with what researchers call physics-informed AI. Their system, known as Magnetic Resonance Artificial Intelligence Velocimetry, or MR-AIV, uses machine learning and established laws of physics to estimate fluid velocity, pressure and permeability from MRI data. The team published its findings on May 27, 2026, in Science Advances.
The technology addresses one of the central challenges in studying the glymphatic system. MRI scans can show how an injected tracer spreads through the brain over time, but they cannot directly measure the speed or direction of fluid flow. MR-AIV analyzes those tracer patterns and reconstructs a three-dimensional map of fluid movement throughout the brain, including regions too deep to be observed with existing imaging methods.
The University of Rochester has been at the center of glymphatic research since 2012, when neuroscientist Maiken Nedergaard, co-director of the university’s Center for Translational Neuromedicine, first described the system. That discovery transformed scientists’ understanding of how the brain removes waste and sparked growing interest in the role impaired fluid circulation may play in neurodegenerative disease.
In experiments involving five healthy mice, researchers found consistent patterns of fluid circulation throughout the brain. The AI-generated maps revealed that glymphatic transport operates through two distinct systems.
One consists of relatively fast-moving pathways located in open spaces surrounding major blood vessels and along the brain’s outer surfaces. Fluid in these regions moved at speeds of several microns per second. The second system involves much slower movement through the brain’s deeper tissues, where fluid traveled roughly 50 times more slowly.
“The results showed that there are two main ways that the glymphatic system washes away particles in the brain,” according to the University of Rochester. Researchers found that fast-moving fluid circulates through open spaces around the brain, while much slower flow moves through deeper tissue.
The findings support the theory that the brain relies on both rapid transport routes and slower tissue-level circulation to remove waste products. Researchers also found that fluid movement varied significantly by region. Faster flow was observed near structures such as the Circle of Willis, a ring of arteries at the base of the brain, and in areas surrounding the olfactory bulb. Much slower flow occurred in regions including the hippocampus and thalamus.
The ability to map those pathways is significant because disruptions in glymphatic circulation have been associated with neurological conditions. Scientists believe impaired waste clearance may contribute to the buildup of toxic proteins and other substances linked to neurodegeneration.
Researchers caution that the work remains in its early stages. The current study was limited to mice, and additional research will be needed to determine whether the same techniques can reliably measure fluid circulation in human brains.
Still, the implications are potentially far-reaching. Dynamic contrast-enhanced MRI, the imaging technique used in the study, is already employed in clinical settings, raising the possibility that the AI framework could eventually be adapted for human patients.
“We’re working hard toward being able to measure the flow of water-like fluids in and around human brains because then the clinical applications get a lot more important and exciting,” Kelley said. “We hope to someday be able to see whether an Alzheimer’s patient has poor circulation in their brain or even screen for poor circulation earlier in life to try to stave off Alzheimer’s. Or we could check when somebody has been concussed to see whether the fluid circulation in their brain is disrupted. This study gets us a step closer.”
The work reflects a broader transformation underway in medicine, where AI is increasingly being used not only to analyze data, but also to uncover patterns and relationships that were previously beyond the reach of conventional tools.
At Harvard Medical School, researchers recently developed an AI system known as Dr. CaBot that can reason through complex medical cases and explain its diagnostic process step by step. The system became the first AI tool to have a diagnosis published alongside that of a physician in The New England Journal of Medicine. While Dr. CaBot is designed to assist with diagnosis and medical education, the Rochester research demonstrates a different but equally significant application of AI: helping scientists visualize and measure biological processes that cannot be directly observed.
Together, the efforts illustrate how AI is expanding medicine’s capabilities at multiple levels, from helping clinicians identify disease to helping researchers better understand the underlying mechanisms that cause it.
Globally, at least 55 million people are living with Alzheimer’s disease and other forms of dementia, according to the World Health Organization. As scientists search for new ways to understand, diagnose and ultimately prevent neurological disease, AI is increasingly emerging as one of the most powerful tools available.
