Computational Imaging, Inverse Problems, and Optical Manipulation
This research community develops mathematical algorithms and physical systems to reconstruct images from indirect measurements and to manipulate microscopic objects using light. It bridges the gap between raw sensor data (from ultrasound, radar, or lasers) and usable visual information, while also using focused light to trap and move biological and material particles.
The work centers on solving inverse problems, where physical parameters are estimated from observed data. Key methods include the Radon transform, Fourier analysis, and compressed sensing, applied to photoacoustic tomography, synthetic aperture radar, and seismic imaging. A significant portion of the research applies deep learning and neural networks to accelerate image reconstruction and target detection. Concurrently, the community advances optical trapping and tweezers, using laser beams to manipulate single molecules, cells, and nanoparticles for biophysical measurement and micro-manipulation. These strands converge on the use of wave physics—acoustic, electromagnetic, or optical—to probe and interact with matter at various scales.
The largest share of the community's output is found in radon research, comprising 9.9% of all radon-related papers tracked, with 4,785 papers in this group. Silicon and gold research also appear, though with negligible shares of their respective element totals.
There are 27,144 papers in total, published most frequently in arXiv (Cornell University), Inverse Problems, and Ultrasonic Imaging.
Recent work continues to focus on integrating transparent ultrasound transducers for combined photoacoustic and ultrasound imaging in clinical settings, as well as developing high-precision optical fiber grippers for 3D micromanipulation.