Physicists at ETH Zurich and EPFL have developed a radically new type of particle detector called PLATON (Plenoptic Light-field Advanced Tracking of Neutrinos). It uses a plenoptic camera — the same technology behind Lytro's "light-field" photography — combined with highly sensitive photon sensors and artificial intelligence to reconstruct the paths of elementary particles in three dimensions. Instead of layering thousands of individual detector channels, PLATON captures the entire particle track in a single block of scintillating material, dramatically simplifying the hardware.
The core innovation is the use of a microlens array placed between the scintillator block and the camera sensor. When a charged particle passes through the scintillator, it emits a faint flash of light. The microlenses split that flash into multiple perspectives, allowing the system to calculate the particle's trajectory in 3D from a single image. The AI then reconstructs the full path with sub-millimeter precision. The team published their proof-of-concept in Nature Communications, showing that PLATON can match or surpass the performance of conventional segmented detectors while being far cheaper and easier to scale.
The implications extend beyond fundamental physics. The same technology could be adapted for medical imaging — particularly positron emission tomography (PET) scans, which currently rely on expensive ring-based detectors. A light-field PET camera could be cheaper, faster, and produce sharper 3D images of metabolic activity inside the body. For neutrino physics, PLATON could enable a new generation of compact detectors that fit in a university laboratory instead of requiring kilometer-scale underground tanks. The team is now working on a full-scale prototype capable of detecting actual neutrinos from nuclear reactors or particle accelerators.
Knowledge takeaway: PLATON (ETH Zurich/EPFL, Nature Communications, 2026) uses a plenoptic light-field camera with a microlens array to capture 3D particle tracks in a single block of scintillator material, replacing complex multi-channel detector arrays; AI reconstructs sub-millimeter-precision paths from the light-field data; the same technology could lead to cheaper, higher-resolution PET scanners for medical imaging; the approach could shrink neutrino detectors from kilometer-scale facilities to lab-bench instruments.