Researchers at Duke University have developed a technology for ultrafast functional photoacoustic
photoacoustic microscopy, whichused by scientists, is based on a combination of light and sound. The laser directs light into the target tissue or cell. Under the action of light, the cell heats up and instantly expands, creating an ultrasonic wave, which is recorded by sensors.
Image: Xiaoyi Zhu et al., Nature Light: Science & Applications
The researchers improved the device for photoacoustic microscopy, and also applied machine learning to improve the quality of incoming images.
Multi-angle scanning system created byscientists, sends more laser pulses over a larger area, and a new scanning mechanism allows the laser scanner and ultrasonic sensor to work simultaneously. The developers note that these changes have doubled the speed of their device.
At the second stage, scientists created an algorithm for machinetraining, which increased the resolution of images. The researchers trained AI to identify the vasculature in the brain using more than 400 images of mouse brains collected in previous experiments. While each brain is unique, the algorithm learned to identify common structures and used that knowledge to fill in previously missing pixels.
"The resulting images looked the sameas detailed as the high-resolution images we would normally get if we ran at much slower speeds and didn't have to sacrifice the full field of view,” says Junjie Yao, one of the study's authors.
The researchers note that when visualizingThe brain needs to do two things at the same time. On the one hand, the instruments must be fast enough to capture short-term events, such as the firing of a neuron or the movement of blood through a capillary. And at the same time, they should show activity on different scales - in the entire brain or at the level of one artery.
You can achieve these goals individually, butit's very difficult to do them all together. It's like choosing between a small, fast car that's uncomfortable to sit in, or a big, spacious car that doesn't go over 30 miles per hour.
Junjie Yao, assistant professor of biomedical engineering at Duke University, co-author of the study
Researchers plan to use UFF-PAM tostudying brain diseases such as dementia, Alzheimer's or even long-term COVID. They also plan to improve the device for imaging organs such as the heart, liver and placenta, which are also in motion and require high speed imaging.
Compare how the lunar eclipse was filmed by NASA and Roscosmos
You have never seen the Sun so close. Solar Orbiter video published
Voyager 1 is sending strange data. NASA doesn't understand what's going on