UC Berkeley computational microscope hits 25.2 billion pixels per second in Nature Photonics study

A UC Berkeley-led team built a computational microscope that captures micron-scale resolution across multi-centimeter areas at faster-than-video rates, Phys.org reported on a Nature Photonics study (DOI: 10.1038/s41566-026-01974-4). The system uses an array of 48 camera sensors with a custom diffractive phase mask and reconstruction algorithms. Principal investigator Laura Waller said the team achieved 3-micron resolution across 5 square centimeters at 120 frames per second. Phys.org and the paper summary state the setup captured 25.2 billion pixels per second. The sensor array fits on a circuit board about the size of a credit card and is treated as a single giant sensor, with the phase mask redirecting light that would otherwise fall between sensors. Lead author Kevin C. Zhou said the combination of high resolution, wide field of view and high speed allowed tracking of freely moving C. elegans and some structures within them. Waller said the approach could image many live organisms simultaneously and monitor samples over time.