Mini microscope enables real-time 3D brain imaging in freely moving mice

Researchers from the University of California in Davis have created a miniaturized microscope for non -invasive imaging in real time, high resolution and non -invasive of brain activity in mice. The apparatus is an important step towards the revolution of the way neuroscientists study the brain.

“What we are doing is to create a technology for image brain activity in displaced mice and behaving freely to open the behavior paradigm,” said Weijian Yang, professor of electricity and computer engineering. “The objective is to create a device capable of allowing research on brain activity and behavior in mice in real time – to see how brain activity stimulates behavior or perception.”

The microscope will advance information on the functioning of the brain, which should benefit human health by strengthening the development of new and improved therapeutic strategies for brain disorders.

The first kind of imagery system, known as DeepinMiniscope, is described in an article published on September 12 Scientific advances.

Iterative design

DeepinMinisCope is based on previous work of Yang to create a lens without lens capable of producing three -dimensional images from a single exhibition.

This imaging system was well suited for large objects in environments with a minimum diffusion of light, such as a robotic vision for the assembly of parts, but had difficulty capturing the details of the biological or biomedical samples. In living tissues, the diffusion of light is widespread, the contrast of the signal tends to be low and to reconstruct the complex characteristics through a large volume of space is a calculation problem.

DeepinMinisCope solves these problems with a new mask design containing more than 100 miniaturized high resolution objectives. A new neural network combines images of each lens to rebuild 3D images.

Deep information (learning)

The DeepinMinCope neural network combines different approaches to automatic learning in order to create a unrolled neural network, which allows instant, precise and high resolution of fine details through a large 3D volume. Using this tool, Yang and his research team recorded the neuronal activity of a mouse in real time.

“Our algorithm combines interpretability, efficiency, scalability and precision,” said Feng Tian, ​​a postdoctoral researcher in the Yang laboratory and first author on the corresponding article. “This requires only a minimum quantity of training data, but it can process in a robust manner and precisely large -scale data sets at high speed.”

Hat trick

By making its microscope small and ergonomic enough for a mouse to carry comfortably and safely when it moves freely, Yang aims to allow neuroscientists to study behavior in real time.

Only 3 square centimeters, at the size of a grape and around the weight of four cents at 10 grams, DeepinMinisCope is almost there.

When earlier, similar conceptions were limited by the large imprint of a traditional camera, DeepinMinisCope uses such a compact sensor as a naked printed circuit with an image sensor, rather than an autonomous and closed system.

Yang’s ultimate goal is a device of 2 square centimeters, which he compares to the size of a hat for a mouse. In addition, for the next iteration, Yang wants to make the device wireless.

“By allowing real -time observation of brain activity in mice that behaves freely, this technology not only increases our fundamental understanding of the way the brain processes information and stimulates behavior, but also contributes to improving our understanding of brain disorders and the development of future therapeutic strategies in humans.”