Rockefeller University captures brain activity of mice on the move

02 Jul 2024

New design of miniaturized head-mounted mesoscope enhances views of neuronal activity.

A project at the Rockefeller University has designed a new lightweight microscope to capture the large-scale brain activity of freely moving mice.

Published in Nature Biomedical Engineering, the findings give researchers new insights into the relationship between neuronal dynamics and animal behavior.

Mounting a functioning imaging microscope onto the heads of live mice is a design challenge that researchers have in the past tackled in several different ways.

Examples have included the MINI2P probe, in which the optical components needed for two-photon imaging were reduced in size through specialized stacked MEMS tunable piezo-membrane lenses; and the use of structured illumination to enhance the focal depth of fluorescence microscopy without requiring any moving parts.

But these approaches have remained flawed, according to Alipasha Vaziri of Rockefeller's Laboratory of Neurotechnology and Biophysics.

"In recent years we’ve seen an explosion of head-mounted microscopes for mice, but they typically support only imaging fields of view of a few hundred micrometers at cellular resolution, since the involved design complexity for larger fields of view comes with an unsustainable weight penalty," said Vaziri.

The core problem, according to the Rockefeller project, has been designs geared toward making whatever technology already existed weigh less while still maintaining the basic optical design of microscopes, in particular the use of a relatively heavy lens in the optical architecture.

Rockefeller took the reverse approach, by first defining what the goals of the instrument were - solving a high-resolution mapping problem between points in a 3D volume of the sample to points on the 2D surface of a camera - and then setting out to create a lightweight system that met those goals, without feeling constrained by the need to adapt a lens-based system.

Fresh thinking for a lens-free strategy

The team's answer is based around diffractive optical elements (DOE), in which microstructures on optical surfaces deflect light at precise predetermined angles. These have found multiple industrial uses, such as the laser machining platform developed at Laser Center Hanover in which DOEs allow a laser to hit multiple locations on a material and machine multiple bores simultaneously.

Using DOEs, the Vaziri lab demonstrated that it is possible to map the positions between a scene and a sensor accurately without forming an image, and then use computational methods to reconstruct the original scene; all without the use of a hefty compound lens to weigh the instrument down.

In trials, Rockefeller's 2.5-gram device was attached to mice subjects and captured broad sections of the mouse brain across a 3.6 x 3.6 square millimeter field of view. It achieved 4-micron lateral resolution, 300-micron depth of field, and recording speed of 16 volumes per second.

Since most of the instrument's parts can be 3D printed or make use of inexpensive, consumer-grade cell phone camera sensors, equivalent modules should be feasible at low cost for any research group interested in testing the technology.

The project next plans to investigate wireless data links for the device, which at present uses physical cables; not a huge problem for use with a single mouse, but the team would like to observe multiple mice interacting with one another.

"The system comes with some sacrifices, and is not nearly as high-performance as larger microscopes," said Alipasha Vaziri. "But this is a key innovation, and one that could only have come from bringing fresh thinking to the problem and freeing oneself from perceived constraints."