UC Berkeley team builds microscope capturing 25 billion pixels per second across 5 cm²

Researchers led by UC Berkeley have developed a computational microscope that simultaneously achieves high resolution, a wide field of view, and fast imaging speeds, a combination that has long eluded optical engineers. The system captures 3-micron resolution across 5 square centimeters at 120 frames per second, producing a data rate of 25.2 billion pixels per second. The study describing the instrument was published in Nature Photonics.

The microscope uses a prefabricated array of 48 disjoint camera sensors mounted on a single circuit board roughly the size of a credit card. Together the sensors act as one giant detector, but gaps between them would normally lose light and data. To recover that information, the team fabricated a custom diffractive optical element — a glass phase mask — that redirects light from the gap regions onto the active sensor areas. A computational algorithm then reconstructs the full image from these compressed measurements.

This hardware-software co-design eliminates the need for laborious manual calibration typically required for large-scale microscope arrays. Co-author Chaoying Gu, a Ph.D. student in electrical engineering and computer sciences, noted that calibration-free operation will be a key ingredient for scaling up future imaging systems.

The researchers demonstrated the microscope on both static and dynamic samples. For static specimens, global features matched low-resolution traditional images while delivering higher resolution detail. For dynamic imaging, they recorded dozens of freely moving C. elegans nematodes at 120 frames per second for 15 seconds, resolving individual worms and internal structures such as the rapidly pumping pharynx involved in feeding.

Lead author Kevin C. Zhou, formerly a postdoctoral researcher in Waller's lab and now an assistant professor at the University of Michigan, said the simultaneous capabilities allowed tracking of individual organisms and functional imaging of fast biological processes. Principal investigator Laura Waller, professor of electrical engineering and computer sciences, described the work as a breakthrough in computational microscopy, achieving the largest space-bandwidth-time product of any practical microscope known to the team.

Collaborators on the study included researchers from UCSF, the University of Utah, UC San Diego, Duke University, and Ramona Optics. Funding came from the Office of Naval Research, the Air Force Office of Scientific Research, the National Institutes of Health, and the Japan Society for the Promotion of Science. Waller is also a Chan Zuckerberg Biohub SF investigator.

The team suggests the approach could enable imaging many live organisms simultaneously and monitoring samples over extended periods. Waller emphasized that joint design of hardware and software has significant potential for scaling up large-scale video microscopy.