HalftoneBalanced · AI Synthesis
technology

New Camera Tracks Invisible Particles in 3D

Scientists developed a new particle detector, PLATON, that uses a light-field camera and AI to track invisible particles in 3D, potentially improving medical scans.

AI-SynthesizedJuly 18, 20261 min read
New Camera Tracks Invisible Particles in 3D

Scientists have developed a new particle detector, called PLATON, that can track invisible particles in three dimensions. This technology could replace millions of tiny detector components with a single block of light-producing material. The system uses a light-field camera, sensitive photon sensors, and artificial intelligence (AI) to reconstruct particle paths. It offers fast, detailed 3D imaging.

PLATON’s design may match or surpass current detectors while being easier to scale. The technology could also lead to sharper medical scans, such as Positron Emission Tomography (PET). Researchers at ETH Zurich and EPFL developed the prototype. Their work was recently published in *Nature Communications*.

Traditional particle detectors often divide scintillating material into many small sections. Optical fibers then collect light from each section and send it to sensors. This method is precise but complex and expensive to scale for larger detectors. PLATON takes a different approach by using advanced camera technology to reconstruct where light originates within an unsegmented block of scintillator material.

The prototype combines a micro-lens array with a single-photon avalanche diode (SPAD) imaging sensor. This allows the system to capture both the intensity and direction of incoming light. The SPAD sensor, developed by EPFL, also provides gated photon detection. This feature helps filter out background noise by recording photons only within specific time windows.

Simulations suggest that an unsegmented PLATON detector could achieve sub-millimeter spatial resolution. The system uses a neural network, adapted from large language models, to process images. This AI analyzes patterns in recorded photons to reconstruct particle interactions. The researchers are developing an upgraded SPAD array sensor to improve photon detection efficiency and provide sub-nanosecond timing for individual photons.

Beyond particle physics, the technology has potential applications in medical imaging. The ETH Zurich researchers have filed three patents for PLATON technology in PET scans. These patents cover both scanner design and image-processing techniques, including the developed neural network. This innovation could improve the imaging of radioactive tracers within the body.

Keep reading

Related stories