A 1-Gram Wireless Microscope Records 8,500 Brain Cells at Once in Freely Moving Mice

To look into the brain of a living animal, researchers usually mount a miniature microscope on its head and connect it with a cable. The cable limits the animal's movement, and existing devices have a narrow field of view, so they can see only a small number of neurons at a time. A team led by Qionghai Dai at Tsinghua University in China reported in Nature a wireless mesoscope called "Airscope" that greatly widens these limits.

A shorter aspheric lens and wireless transmission

Airscope weighs 1.0 g and has a lateral resolution of 4 μm, a field of view 6 mm in diameter and a frame rate of 12.5 frames per second. Its aspheric lens design makes the optical path 2.84 times shorter than in earlier miniature microscopes, yet it captured about as many neurons as a single-photon mesoscope fixed to a bench. With a dedicated wireless transmitter, it recorded calcium signals from more than 8,500 neurons per mouse without cables, which the researchers described as a 25-fold increase over earlier devices.

The researchers tested the device in a room-sized play area, in a swimming task and in simultaneous recording of several animals. When two mice competed, activity patterns across several areas of the dorsal cerebral cortex were linked to social rank, and group activity could predict the outcome of the competition in advance. During maze navigation, cortical activity tied to the reward location appeared, and recording the brain and spinal cord at the same time let the team analyze the flow of signals from the spinal cord to the cortex (Granger causality) during fear conditioning.

Significance and limits

Recording the whole cortex at cellular resolution in naturally behaving animals has long been a challenge in systems neuroscience. This device could make such experiments comparatively easier. However, this is a tool-development study in mice, and because of the nature of single-photon calcium imaging, neurons in deep layers are hard to see and the time resolution falls short of electrophysiological recording. A correlation between behavior and brain activity is not necessarily causation.