For decades, the very instruments designed to illuminate the brain have obscured what they sought to reveal — their noise stressing the animals under study, their artifacts blurring the signals researchers chased. A team at the University of North Carolina at Chapel Hill has developed SORDINO, a quieted fMRI technique that images freely moving, unanesthetized mice with clarity, removing two of the field's most persistent confounds at once. Named for the musical instruction to mute, it arrives as both a technical achievement and a reminder that the most elegant solutions often come not from inv
Silent fMRI technique enables brain imaging in moving, behaving mice
The birth of a new approach—people will have to play with it
So the main problem SORDINO solves is that conventional scanners are too loud for awake animals?
That's part of it, but it's deeper. The noise stresses the animals, which changes their brain activity. And if you use anesthesia to avoid the stress, you're measuring a drugged brain, not a normal one. SORDINO lets you image awake, behaving animals without either problem.
But we should be clear—this is tested in mice. We don't know yet if it works in humans or in weaker scanners.
What makes SORDINO different technically?
Instead of sweeping magnetic fields back and forth like conventional scanners, it rotates smoothly, like clock hands. That keeps the field nearly constant and eliminates the noise-generating switching.
And the signal it detects is different from standard fMRI, right? It's measuring tissue oxygen and blood volume rather than the difference between oxygenated and deoxygenated hemoglobin.
Yes, which could give more localized information. But one expert noted the study didn't completely rule out whether they're detecting dissolved oxygen in blood rather than oxygen in tissue itself.
Why does that distinction matter?
It matters for interpretation. If you're measuring dissolved oxygen, you're getting a different kind of information about what's happening in the brain than if you're measuring oxygen bound to hemoglobin.
Right. And that affects how you understand what the signal actually tells you about neural activity.
What's the practical next step?
Shih's team is trying to adapt it for humans—particularly for people who can't tolerate conventional scanners, like infants or people with autism or PTSD.
But that's still uncertain. The technique worked at high resolution in mice. Human studies use lower resolution, where physiological noise becomes a bigger problem. We don't know if SORDINO will work in that context.
Le Pouls
- Conventional fMRI scanners exceed 100 decibels during operation, stressing animals, corrupting data, and forcing researchers into the distorting compromise of anesthesia.
- SORDINO eliminates both problems by rotating smoothly through magnetic space rather than sweeping, slashing noise to idle levels while still capturing brain activity in mice that move, eat, and interact.
- The technique opens an experimental frontier previously inaccessible: simultaneous imaging of two interacting animals, and pairing with electrophysiology to compare electrical signals with blood-flow changes during social behavior.
- Its signal differs subtly from standard BOLD contrast, appearing sensitive to tissue oxygenation itself — a finer, more localized measure that excites researchers even as questions about its precise physiological source remain open.
- Shih's team is now adapting SORDINO for human use, targeting infants and patients with autism, PTSD, or movement disorders — populations for whom the silence of the scanner may matter as much as its resolution.
For decades, the very instruments designed to illuminate the brain have obscured what they sought to reveal — their noise stressing the animals under study, their artifacts blurring the signals researchers chased. A team at the University of North Carolina at Chapel Hill has developed SORDINO, a quieted fMRI technique that images freely moving, unanesthetized mice with clarity, removing two of the field's most persistent confounds at once. Named for the musical instruction to mute, it arrives as both a technical achievement and a reminder that the most elegant solutions often come not from invention but from recombination — familiar principles arranged in a new key.
Scientists trying to image awake animal brains have long been caught between two bad options: scanners loud enough to stress and confound the very subjects under study, or anesthesia that quiets the animals while distorting the neural activity researchers are trying to see. A team led by Yen-Yu Ian Shih at the University of North Carolina at Chapel Hill has built a way out of that bind.
Their technique, SORDINO — a musical term for muting — works by rotating smoothly through magnetic space rather than sweeping back and forth, keeping the field gradient nearly constant and compressing the gap between radiofrequency pulses and data collection to almost nothing. The result is a scanner that operates at idle-level noise while producing clear images of freely moving, unanesthetized mice. Published in Nature Neuroscience, the work demonstrated the method across several scenarios: resting brain activity, the neural patterns of mice reaching for food, and the simultaneous imaging of two animals interacting with each other.
The physics behind SORDINO is not new — it draws on principles already embedded in existing MRI methods. What distinguishes it is how those principles have been combined and applied to functional imaging. Ravi Menon, a medical biophysicist at the University of Western Ontario, called it a clever recombination of established ideas, and contrasted it favorably with a recent controversial technique that was retracted. The signal SORDINO captures also differs subtly from standard BOLD contrast, appearing sensitive to tissue oxygenation itself — a more localized measure that could offer finer anatomical detail, though questions about its precise physiological source remain.
Shih's group is already working to bring SORDINO to human scanning, with particular attention to populations who struggle with conventional machines: infants, and people with autism, PTSD, or movement disorders. Whether the technique holds up at the lower resolutions typical of human studies — where breathing and heartbeat noise become more disruptive — is the central challenge ahead. For now, many labs at Shih's institution have adopted it for animal work, and the broader field is watching to see whether it travels well to other scanners and other hands.
Scientists have long faced a stubborn problem when trying to image the brains of awake animals: the machines themselves get in the way. A conventional fMRI scanner produces noise exceeding 100 decibels as it sweeps magnetic fields back and forth, a sound so loud it stresses the animals being studied and creates artifacts that blur the very brain activity researchers are trying to measure. Anesthesia solves the noise problem but introduces a different one—sedatives alter brain function in ways that make it impossible to know whether you're seeing the animal's actual neural activity or the drug's effects.
A team led by Yen-Yu Ian Shih, a neurology professor at the University of North Carolina at Chapel Hill, has developed a technique that sidesteps both problems. Called SORDINO—a musical term meaning to mute—the method captures brain images by rotating smoothly through space rather than sweeping back and forth, keeping the magnetic field gradient nearly constant while reducing the gap between radiofrequency pulses and data collection to almost nothing. The result, published last month in Nature Neuroscience, is a scanner that operates at idle-level noise while still producing clear images of moving, behaving animals.
Shih's team tested SORDINO on awake mice in several scenarios: measuring their resting brain activity, mapping the neural patterns that accompany reaching for food, and simultaneously imaging the brains of two mice interacting with each other. In each case, the technique captured usable data even as the animals moved freely. The noise reduction also made the scanner compatible with other recording methods, particularly electrophysiology, which measures electrical activity directly from neurons—a pairing that could reveal how electrical signals and blood flow changes relate to one another during social behavior.
The physics underlying SORDINO is not new; it draws on well-established principles already embedded in many MRI techniques. What makes it novel is how those principles have been combined and applied to functional imaging. Unlike a controversial technique called DIANA that was eventually retracted, SORDINO rests on solid physiological ground, according to Ravi Menon, a medical biophysicist at the University of Western Ontario who was not involved in the work. "They've combined a bunch of existing physical principles into a very clever application," Menon says.
The signal SORDINO detects differs subtly from the standard BOLD (blood-oxygen-level dependent) contrast used in most fMRI studies. While both methods rely on the relationship between neural activity and blood flow, BOLD measures the difference between oxygenated and deoxygenated hemoglobin, whereas SORDINO appears sensitive to cerebral blood volume, blood flow, and tissue oxygenation itself. That last capability intrigues researchers because tissue oxygen signals are more localized than BOLD signals, potentially offering finer anatomical detail. Peter Bandettini, chief of functional imaging methods at the National Institute of Mental Health, cautioned that the study did not definitively rule out whether the oxygen signal comes from dissolved oxygen in blood rather than oxygen bound to hemoglobin—a distinction that matters for interpretation. Still, he found the paper convincing and sees potential for human applications.
Shih's group is already working to adapt SORDINO for human scanning, with an eye toward populations who struggle with conventional scanners: infants, people with autism, those with PTSD or movement disorders. The challenge will be whether the technique maintains its effectiveness at the lower spatial resolutions typically used in human studies, where physiological noise from breathing and heartbeats becomes more problematic. At the high resolution used in the mouse experiments, such noise is minimal.
For now, Shih's team and many other researchers at his institution have largely switched to SORDINO for animal fMRI work. But questions remain about how broadly the method will travel. Menon and Alessandro Gozzi, director of the Functional Neuroimaging Laboratory at the Istituto Italiano di Tecnologia, both noted that success will depend on whether other labs can implement it reliably and whether it works in less-powerful scanners. "This is the birth of a new approach," Menon says. "People will have to play with it and see if it works."
Citations marquantes
The degree to which the new method reduces motion artifacts is really remarkable. The fact that it is silent is just a bonus.— Alessandro Gozzi, director of the Functional Neuroimaging Laboratory at the Istituto Italiano di Tecnologia
They've combined a bunch of existing physical principles into a very clever application for functional MRI.— Ravi Menon, professor of medical biophysics at the University of Western Ontario