Summary
A bioRxiv preprint describes COMET, a 3.75-g head-mounted imaging system that recorded more than 15,000 cortical neurons in freely moving mice during five-minute sessions. The researchers used it to examine brain activity during social interaction and fear conditioning.
A bioRxiv preprint describes COMET, a 3.75-g head-mounted imaging system that the authors say can record calcium activity from more than 15,000 neurons in the outer layers of the mouse cortex during a five-minute session. Its field of view can reach 10 mm in diameter, and the researchers used it to study freely moving mice during social interaction and an auditory trace fear-conditioning task. The preprint was posted on October 15, 2026.
How COMET combines a wide view with cellular detail
Studying brain activity during natural movement is difficult because a small field of view can miss coordination across distant regions, while expanding the view can make it harder to distinguish individual neurons. A head-mounted instrument also has to be light enough for the animal to move while providing a usable signal.
COMET—short for Cortex-wide Observational Miniature Epifluorescence Technique—combines an aberration-controlled wide-field optical design with uniform, high-efficiency illumination. The authors also describe a single coaxial connection for power and data, alongside a processing pipeline for denoising images and extracting neural signals. Calcium imaging detects fluorescence changes associated with calcium inside cells; researchers use those changes as an indirect readout of neuronal activity.
The authors report that COMET resolved individual cells across broad cortical fields while mice moved freely. In comparisons with benchtop two-photon microscopy, they report high spatial correspondence between detected neurons and similar calcium dynamics at the level of individual events.
Testing the system during behaviour
The researchers applied COMET to social interaction, using it to quantify how different brain regions engaged. They also recorded during auditory trace fear conditioning, a task involving natural movement and behavioural expression. In that experiment, the authors report time-ordered changes in regional firing-rate patterns and functional connectivity before a global conditioned response.
The value of this approach is its combination of single-cell detail and a wide cortical view during behaviour. It may help researchers investigate how activity in distributed brain regions changes together as an animal responds to its surroundings, rather than studying those signals only under more constrained conditions. The reported experiments demonstrate the system in mice; they are not evidence of a human application.