Summary: Multidisciplinary neuroengineering research has broken a major imaging barrier by mapping the precise flow velocity of the brain's garbage disposal infrastructure. The study uses artificial intelligence known as physics to understand magnetic resonance imaging (MRI) data, which reveals the hidden mechanisms of the glymphatic system, a fluid network that washes away metabolic waste such as amyloid-beta proteins linked to Alzheimer's disease.
AI models discovered a double-speed drainage pattern, showing that the protective fluid moves 50 times faster in the outer parts of the brain than it does when it flows through the deep brain tissue.
Important Facts
- Glymphatic Baseline: First described in 2012 by pioneering neuroscientist Maiken Nedergaard, the glymphatic system acts as the brain's internal network of pipes. When the subject enters deep sleep, a fluid like water circulates around the central nervous system to clean up the metabolic waste associated with neurodegenerative diseases.
- MRI Velocity Limitation: Examining this internal fluid circulation within the living brain has historically been almost impossible. Conventional microscopes provide high-detail views of a small piece of tissue, while traditional 3D MRIs do not have the sensing capability to track fluid flow velocities in slow motion.
- Physics-Informed Neural Networks: To bridge this technology gap, mechanical engineers and computer scientists have developed physics-based AI tools. By training neural networks on MRI videos to track the dye across brain tissue over time, the AI successfully detected fluid flow velocities and mapped tissue infiltration.
- The Dual-Velocity Blueprint: AI revealed that the glymphatic system clears toxic amyloid-beta particles using two different speeds. The “fast track” moves fluid at a few microns per second through open cortical areas, such as the space between the skull and the brain. In contrast, the “slow track” flows through deep brain tissue at a rate 50 times slower.
- Diagnostic Roadmap: Although current basic measurements have been successfully established in animal models, researchers are developing AI software for human clinical settings. The goal is to compare fluid dynamics across the young, old, healthy, and sick.
- Preventing Trauma and Dementia: Senior author Professor Douglas Kelley notes that mastering this fluid mapping brings science closer to diagnosing childhood brain abnormalities to prevent Alzheimer's disease. Additionally, technology can be deployed immediately after a concussion to assess whether a patient's internal fluid circulation has been dangerously disrupted.
Source: University of Rochester
When a person sleeps deeply, a fluid like water surrounds the brain, cleaning out the wastes that are processed in the body and related to diseases such as Alzheimer's.
This process, known as the glymphatic system, was first described in 2012 by Maiken Nedergaard—a pioneering neuroscientist and director of the University of Rochester Center for Translational Neuromedicine.
But questions remain about the mechanics of the system—in particular, how quickly fluid circulates in the brain. Studying the circuitry of the living brain is difficult to do without causing irreversible damage to the subject.
“You can put a microscope on a small part of the brain and look at what's going on there in great detail, and we've worked with that kind of data in the past, but it's a small view of the whole process,” said Professor Douglas Kelley of the Department of Mechanical Engineering at URochester.
“If you want to image the entire brain, MRI is a great option because it gives you a three-dimensional view. But MRI has serious limitations as well, the biggest of which is that it doesn't capture the velocity of fluid flow, at least not because of that much flow.”
Kelley and colleagues at Rochester, Brown University, and the University of Copenhagen turned to artificial intelligence for help.
In a new study published in Advances in Sciencedescribe how they used physics-informed artificial intelligence to determine fluid flow velocities from magnetic resonance imaging (MRI) data. Using videos of dyes spreading through brain tissue over time, the neural networks the researchers built were able to determine how fast the fluid flows and how permeable the brain tissue is.
The results showed that there are two main ways that the glymphatic system washes away particles in the brain like amyloid beta proteins that are linked to Alzheimer's disease—and one of these ways is faster than the other. The rapid flow of water-like fluid of the glymphatic system moves at a few microns per second in the open area of the brain such as the space between the skull and the brain, while the slow flow of water-like fluid flows through the deeper tissues of the brain at a rate 50 times slower.
Until now, researchers have been working to find basic measurements of fluid flow in the brains of animals such as mice to inform AI tools. In the future, they hope to be able to compare fluid flow in healthy and diseased brains and young and old brains, with aspirations of eventually studying human circulation.
“We're working hard to be able to measure the flow of fluids like water in and around the human brain because that's where the clinical applications become more important and more exciting,” said Kelley. “We hope that one day we will be able to see if a patient with Alzheimer's disease does not have good blood flow in their brain or check if blood flow is not good early in life to try to prevent Alzheimer's.”
Sponsorship: The research is supported by the NIH National Center for Complementary and Integrative Health and the NIH BRAIN Initiative.
Kelley's research collaborators include Brown University PhD student Juan Diego Toscano, Rochester computational scientist Yisen Guo, Brown University PhD student Zhibo Wang, Rochester PhD student Mohammad Vaezi, University of Copenhagen Associate Professor Yuki Mori, Brown University Professor George Karniadakis, and Rochester Assistant Professor.
Important Questions Answered:
A: It is the brain's unique way of removing biological waste. Like any other functioning organ, the brain produces metabolic waste as it functions throughout the day. During deep sleep, the glymphatic system opens up, allowing fluids like water to circulate through your nervous system and remove toxic substances such as amyloid-beta proteins, which are directly linked to the development of Alzheimer's disease.
A: Because conventional MRIs don't detect fluid movements this amazingly slow. Although a conventional MRI can provide a flawless 3D image of the brain's shape, it lacks the technical capability to quantify the velocity of the glymphatic trickle. By building custom physics-savvy AI, researchers could analyze videos of the dye diffusing through the tissue and determine statistically how fast the fluid flows and how permeable the brain structure is.
A: Provides a quick, non-invasive look at internal structural trauma. When someone is hit hard, the physical impact can disrupt the delicate pathways of the glymphatic system, stopping the brain's ability to clean itself. Using this AI-driven fluid tracking technology, doctors can finally assess a depressed patient to see if their fluid circulation has been compromised, allowing for more intelligent, individualized recovery procedures.
Editor's Notes:
- This article was edited by a Neuroscience News editor.
- The journal paper is fully revised.
- Additional content added by our staff.
About this neuroscience and AI news
Author: Luke Auburn
Source: University of Rochester
Contact person: Luke Auburn – University of Rochester
Image: Image posted in Neuroscience News
Actual research: Open access.
“MR-AIV reveals in vivo large-scale fluid flow in the brain with physics-informed AI” by Juan Diego Toscano, Yisen Guo, Zhibo Wang, Mohammad Vaezi, Yuki Mori, George Em Karniadakis, Kimberly AS Boster, and Douglas H. Kelley. Advances in Science
DOI:10.1126/sciadv.aeb0404
Abstract
MR-AIV reveals extensive cerebral fluid flow in vivo with physics-savvy AI
Cerebrospinal and interstitial fluid circulation plays an important role in removing metabolic waste from the brain, and its disruption has been linked to neurological disorders. However, directly measuring the transport of brain-enriched fluid, especially in the deep brain, has never been possible.
Here, we present magnetic resonance artificial intelligence velocimetry (MR-AIV), a framework containing special physics knowledge of structures and an optimization method that reconstructs three-dimensional fluid velocity fields from dynamic magnetic resonance imaging (DCE-MRI). MR-AIV reveals extensive velocity maps of the brain while providing measurements of tissue permeability and pressure fields, values unattainable by other methods.
Applied to the brain, MR-AIV reveals the functional landscape of interstitial and perivascular flow, quantitatively distinguishing diffusion-driven transport. [∼0.1 micrometers per second (μm/s)] from fast advective flow (∼3 μm/s).
This approach allows for new investigations into brain displacement mechanisms and fluid dynamics in health and disease, with possible wide application to other medium systems, from geophysics to tissue mechanics.
