Spectral DiffuserScope delivers 15 cm hyperspectral snapshot imaging via compressed sensing on standard microscopes

The Spectral DiffuserScope attaches to existing fluorescence microscopes and replaces scanning or bulky filter wheels with a diffuser and computational reconstruction. Researchers from UC Berkeley and collaborating institutions tested the device on fluorescent beads, labeled mammalian cells, and lanthanide-doped hydrogels. Simulations benchmarked it against earlier snapshot methods, showing improved resolution under photon-limited conditions typical of live-cell work. The approach exploits compressed sensing to recover dozens of spectral channels from a single exposure, addressing the throughput limits of sequential hyperspectral systems.

Conventional hyperspectral microscopes either scan wavelengths or use large optics that restrict temporal resolution and field-of-view compatibility. This work builds on earlier diffuser-based cameras and coded-aperture techniques while shrinking the form factor enough for routine lab use. It connects to broader trends in computational imaging that trade hardware complexity for post-processing, similar to advances in light-field and Fourier ptychographic microscopy. If reconstruction artifacts remain low across varied fluorophores, the design could enable routine chemical mapping in materials labs without dedicated core-facility equipment.

Key limitation is reliance on simulation benchmarks plus a narrow set of test samples; real-world performance under autofluorescence or dense labeling remains unquantified. Validation would require side-by-side comparison against a commercial linescan system on the same biological replicates with quantitative unmixing metrics. Next steps include integration with microfluidic devices for higher-throughput screening and open-source reconstruction code to allow community testing.