AI tech helps researchers peer into mouse brains Artificial intelligence and robotics news

Johns Hopkins biomedical engineers have developed an artificial intelligence (AI) training strategy to capture images of mouse brain cells in action. The researchers say the AI ​​system, in concert with specialized ultra-small microscopes, can find precisely where and when cells are activated during movement, learning and memory. The data collected through this technology could one day allow scientists to understand how the brain works and is affected by disease.

The researcher’s experiments on mice were published in Nature Communication March 22.

“When a mouse’s head is retained for imaging, its brain activity may not truly represent its neurological function,” says Xingde Li, Ph.D., professor of biomedical engineering at Johns Hopkins University School of Medicine. “To map the brain circuits that control daily functions in mammals, we need to see precisely what is happening between individual brain cells and their connections, as the animal moves freely, eats and socializes. »

To collect this extremely detailed data, Li’s team developed ultra-small microscopes that mice can wear on top of their heads. Measuring a few millimeters in diameter, the size of these microscopes limits the imaging technology they can carry. Compared to benchtop models, the frame rate of miniature microscopes is low, making them susceptible to interference from motion. Perturbations such as respiration or mouse heart rate would affect the accuracy of the data these microscopes can capture. The researchers estimate that Li’s miniature microscope would need to exceed 20 frames per second to eliminate all motion disturbances from a freely moving mouse.

“There are two ways to increase the frame rate,” Li explains. “You can increase the scanning speed and you can reduce the number of dots scanned. »

In previous research, Li’s engineering team quickly discovered that it hit the physical limits of the scanner, hitting six frames per second, which maintained excellent image quality but was well below the required rate. So, the team moved on to the second strategy to increase the frame rate – decreasing the number of points scanned. However, similar to reducing the number of pixels in an image, this strategy would cause the microscope to capture lower resolution data.

Li speculated that an AI program could be trained to recognize and restore missing dots, thereby enhancing images to higher resolution. Such AI training protocols are used when it is impractical or time-consuming to create a computer program for a task, such as reliably recognizing a cluster of features like a human face. Instead, computer scientists take the approach of letting computers learn to program themselves by processing large sets of data.

A significant challenge in the proposed AI approach was the lack of similar images of mouse brains to train the AI. To fill this gap, the team developed a two-step training strategy. The researchers began training the AI ​​to identify the building blocks of the brain from images of still samples of mouse brain tissue. They then trained the AI ​​to recognize these building blocks in a restrained-headed living mouse under their ultra-small microscope. This step trained the AI ​​to recognize brain cells with natural structural variation and small movement caused by the movement of mouse breathing and heartbeat.

“The hope was that whenever we collect data from a moving mouse, it will always be similar enough for the AI ​​network to recognize,” Li says.

Next, the researchers tested the AI ​​program to see if it could accurately improve images of the mouse brain by gradually increasing the frame rate. Using a reference image, the researchers reduced the microscope scan points by factors of 2, 4, 8, 16 and 32 and observed how precisely AI could improve the image and restore the image resolution.

The researchers found that the AI ​​could adequately restore image quality at up to 26 frames per second.

The team then tested the performance of the AI ​​tool in combination with a mini microscope attached to the head of a moving mouse. Using the combination of AI and the microscope, the researchers were able to accurately see the activity spikes of individual brain cells activated by the mouse walking, turning, and generally exploring its surroundings.

“We could never have seen this information at such high resolution and frame rate before,” Li says. action at the cellular level. »

The researchers say that with more training, the AI ​​program might be able to accurately interpret images at up to 52 or even 104 frames per second.

Other researchers involved in this study include Honghua Guan, Dawei Li, Hyeon-cheol Park, Ang Li, Yungtian Gau, and Dwight Bergles from Johns Hopkins University School of Medicine; Yuanlei Yue and Hui Lu of George Washington University; and Ming-Jun Li of Corning Inc.

This research was supported by the National Cancer Institute (R01 CA153023), the National Science Foundation Major Research Instrumentation Grant (CEBT1430030), and the Johns Hopkins Medicine Discovery Fund Synergy Award.

We would love to give thanks to the writer of this post for this incredible web content

AI tech helps researchers peer into mouse brains Artificial intelligence and robotics news

Take a look at our social media profiles and other pages related to them