A team led by Raju Tomer, professor of biological sciences at Columbia University, has created a new design for microscopes and microscope lenses that could push 3D tissue imaging beyond state-of-the-art systems while drastically cutting costs and complexity. Details of the design were published today in the journal Nature Biotechnology.
Modern biology and medicine increasingly depend on high-resolution, 3D images of intact tissues such as brains and cancer biopsies. The images allow researchers to map neural circuits, characterize disease, and train next-generation AI models for diagnosis, among other uses. But progress in imaging these tissues easily and at a large scale has been bottlenecked by the lenses that capture light from samples. Researchers have had to make tradeoffs: "Oil-immersion" lenses, which touch the sample through a drop of oil, deliver the sharpest images; but they're expensive, can only see a few millimeters deep at most, and require specific sample preparations. Cheaper lenses that work at a distance through air can reach up to several centimeters into a sample, but they produce blurred images when used with the chemicals that render tissues transparent for 3D viewing.
The Tomer team's solution, called HySIL (Hybrid Solid–Liquid Optics), addresses both problems. It pairs a simple, curved solid lens with a precisely matched immersion liquid so the two function as a single continuous optical system. The design lets inexpensive air lenses deliver high-resolution images across centimeter-scale tissues, and across virtually any common sample-preparation method, with no hardware changes.
To demonstrate the framework, the team built a modular device called SCOPE that can be added directly to existing light-sheet microscopes. They also developed an even higher-resolution proof-of-concept variant, called Super-SCOPE.
"We've broken a long-standing trade-off in microscopy between performance and accessibility," Tomer said. "By making the immersion liquid an active optical component rather than a passive filler, we get the resolution of the most expensive lab systems with the cost and footprint of equipment that fits anywhere, from teaching labs to clinics in low-resource settings."
These technologies were also added to a compact, projector-based light-sheet microscope (pLSM) that Tomer's group developed in 2024, which is now available commercially under the name SLICE.
Working with a network of academic collaborators across the life sciences, the paper's authors demonstrated pLSM-SCOPE on whole mouse, salamander, and cavefish brains for mapping neural circuits; lab-grown miniature human brain tissues used to study development and disease; and intact human cancer biopsies for next-generation 3D pathology. Because HySIL is a general framework, it can be attached to other types of microscopes, including confocal, two-photon, and other 3D imaging modalities.
"By making 3D imaging much easier to scale, this work could help fuel the next generation of AI models for disease detection, grading, and prognosis," Tomer said. "None of these demonstrations would have been possible without our wonderful collaborators across neuroscience, developmental biology, and pathology, as well as at our industry partner MBF Bioscience."
Jack Glaser, co-founder and CEO of MBF Bioscience, was a co-author on the paper. "A new optical concept only changes a field if labs without specialized optics expertise can actually use it day to day," Glaser said. "What HySIL gives us is the rare combination of lower cost and higher performance in the same instrument. Engineering that into a robust, well-supported light-sheet system makes it something working labs can rely on across neuroscience, cancer research, developmental biology, and pathology."
For decades, most tissue analysis has relied on thin, two-dimensional slices of tissue placed on glass slides. But many important biological and disease features are better seen in three dimensions, which this new technology allows.
Hanina Hibshoosh, professor of pathology and cell biology at Columbia University Irving Medical Center, was a co-author on the paper.
"Examining 3D images of tissues lets you see the whole tissue architecture, not just the cross-section that pathology has traditionally been limited to," Hibshoosh said. "Tools like pLSM-SCOPE that make this kind of imaging affordable and accessible will become increasingly important as AI helps us analyze ever-larger amounts of tissue data for diagnosis and prognosis."
The research was supported by the National Institutes of Health. Columbia University has filed patent applications related to each of the technologies described in the paper. Tomer has served as a paid consultant for MBF Bioscience for the implementation of pLSM as SLICE, and several of the paper's co-authors are employees of MBF Bioscience.
Others at Columbia who contributed to this work include professors Rene Hen, Christopher Makinson, and Maria Tosches; postdoctoral fellows Sudha Guttikonda and Giacomo Gattoni; doctoral candidates Cheng Gong and Pauline Affatato; and research associate Daniel De La Cruz.
"The kind of access to tissue data that this technology provides is going to be transformative as biology and medicine increasingly depend on dense, tissue-scale image datasets," Tomer said.
