Chinese scientists have developed a wireless microscope weighing only 1 gram that can record neural activity in the mouse cerebral cortex with single-cell resolution in natural environments.

Named Airscope, the device was developed by a team led by Dai Qionghai, an academician of the Chinese Academy of Engineering and a professor at Tsinghua University, in collaboration with researchers from other institutions. Details of the achievement were published in Nature on Monday.

Dai said the microscope provides a new tool for understanding the neural basis of intelligence and will drive optical brain-computer interfaces toward more precise neural detection. Its future applications may expand beyond neuroscience to fields such as immunity and oncology, enabling complete tracking of disease processes, Dai added.

DILEMMA OF CONVENTIONAL METHODS

How the brain regulates behavior in complex natural environments is one of the most fundamental questions in neuroscience and crucial for understanding intelligence and developing brain-computer interfaces. However, such research has long been constrained by a trade-off between the throughput of neural observation achievable in freely behaving animals and the weight of the microscope.

To study how the brain regulates behavior in natural environments, it is necessary to simultaneously record large numbers of neurons across broad brain regions with single-cell resolution while animals can move and interact socially.

Existing microscopes capable of resolving individual neurons, although able to image large areas of the cortex clearly, still require the mouse's head to be fixed under the instrument or tethered with a cable, making it difficult for the animal to move freely and hindering studies of neural and behavioral processes under natural conditions.

At the same time, existing head-mounted microscopes, while enabling free animal movement, usually have a field of view of less than 1 mm, allowing only a small brain region to be seen. To cover a larger cortical area at higher resolution, microscopes become heavier and consume more power.

Solving this dilemma allows researchers to observe how thousands of neurons work together to support behavior in settings closer to real life, opening a new route to understanding the neural basis of intelligence and providing a new tool for large-scale neural-signal readout in optical brain-computer interfaces.

ONE-GRAM MICROSCOPE

Through integrated design and systematic optimization, the Airscope device weighs only 1 gram and can achieve single-cell resolution of about 4 to 6 micrometers over a field of view at 6 mm.

Airscope can be mounted directly on a mouse's head and uses a 2-megapixel image sensor to perform on-device compression, enabling megapixel-level images to be transmitted wirelessly.

Experiments showed mice wearing Airscope could climb ropes and squeeze through tubes with no significant difference in travel distance, speed or path tortuosity, compared with mice not wearing the device.

The quality of neuronal signals recorded by Airscope was comparable to that of conventional benchtop microscopes. Combined with a neuron-extraction algorithm previously developed by the team, Airscope can simultaneously record more than 8,500 neurons across 14 cortical areas in one cerebral hemisphere of a single mouse, about 25 times more than existing wireless imaging devices. It can record at the same site for more than 200 days.

DIVERSE OBSERVATION SCENARIOS

With the wearable device, the team unlocked multiple forms of complex free exploration by mice under natural conditions and completed experiments that were previously difficult due to cables or equipment weight.

These included simultaneously recording cortical activity in multiple mice exploring a 5-square-meter natural habitat, and tracking mice over long periods of autonomous exploration in a large maze with 32 ends.

Standard laboratory cages make it difficult for mice to express natural behaviors such as digging, long-distance exploration and complex social interaction. Instead, the team built a 5-square-meter habitat containing shelters, ramps, tree holes, wooden huts and other terrain features. Here, mice moved more than twice as fast as in cages and displayed 16 categories of behavior, including climbing, going up and down steps, entering holes and crossing different terrain areas.

The study found that cortical activity changed systematically with the terrain area the mouse was in. Using cortical neural activity alone, a deep-learning model could determine which area a mouse was in, with accuracy far above chance.

"Airscope provides greater behavioral freedom for neuroscience research and allows large-scale neural observation to follow animals into richer natural behaviors, helping us better study brain cognition and brain diseases," said Zhang Yuanlong, an assistant professor at Tsinghua University's School of Life Sciences.