Stanford researchers have improved the technology used to diagnose a common form of birth injury in newborns.
Mini-strokes, caused by breaks in tiny blood vessels, can occur during or soon after birth. These brain injuries are fairly common in preemies, though full-term babies can experience them, too. Ultrasound scans of babies’ brains allow doctors to see the injuries, but until now these scans haven’t given much information on tiny patients’ likely long-term outcomes.
The new research expands the capability of ultrasounds beyond showing only brain structure, to provide a real-time window into brain function.
“I think it’s going to be a very useful tool to help us predict the consequences of a brain bleed, and to understand brain plasticity,” said pediatric radiologist Erika Rubesova, MD, a co-author of the new study.
Infants’ brains are good at compensating for injury, Rubesova told me: Young brains have a lot of plasticity, meaning one brain area can often take over the role of another area that’s injured. That’s good news for babies, but it means that doctors aren’t always able to tell new parents what the outcome of a small stroke will be.
Starting with a baseline brain map
The new technology should allow physicians to create detailed maps of normal brain activity in healthy babies, Rubesova said, and then use these maps as a baseline for evaluating the functional impact of small strokes.
The new approach won’t require hospitals to buy new ultrasound machines, either:
“The awesome part is that we used the same ultrasound equipment, and even the data that is acquired is the same as before,” said engineer Marko Jakovljevic, PhD, the study’s lead author. “It’s how you process the data once you get it that is different.”
The new method, described recently in IEEE Transactions on Ultrasonics, Ferroelectrics and Frequency Control, also has a great safety profile — it works on an awake, wiggly baby and requires no radiation.
How does it work?
Ultrasound scanners work by projecting high-frequency sound waves into the body, detecting how these sound waves bounce back from the tissue, and transforming the signals into a picture.
Sound waves don’t penetrate bone, but because newborns have gaps in their skulls — soft spots where their skull bones have not yet joined — it’s possible to point the transducer between the bones to get an image of what’s happening in the brain.
Creating an ultrasound picture requires math, a bunch of equations that “focus” the raw data from bounced-back sound into an image, similar to how a camera’s lenses focus light.
In the new technique, the Stanford team starts with the exact same raw data as before. But they’ve devised a better way to mathematically suppress “motion artifacts,” which are “noisy” aspects of data that come from movement (the baby’s wiggling, the doctor’s hand moving around with the transducer) rather than from what is actually happening inside the brain. Motion artifacts blur the image. Getting rid of these blurs gives a crisper image, and allows the team to see blood flow in real time.
It’s like the difference between a still, long-exposure photo of a highway at night (with streaks of red and white for cars’ lights) and a movie that shows individual cars driving along a highway.
“The method we developed can be used not just for a better image of tiny blood vessels but also for tracking blood flow,” Jakovljevic said.
Better blood-flow tracking
Because greater blood flow is linked to more neural activity, the scan produces a real-time map of brain activity, which has the potential to help physicians assess the impact of a small stroke. “We can ask, ‘Have any parts of the brain been damaged?’ and also, ‘Have any functions of the brain been impaired?'” Jakovljevic said.
The scientists hope the improved ultrasound images will spur new research in how to treat brain injuries.
“I hope it’s not only clinically valuable, but also a scientific tool,” Jakovljevic said, adding, “We think this will help preemies’ long-term health.”
