Sparsity-based reconstruction for super-resolved limited-view photoacoustic computed tomography deep in a scattering medium

Delay-and-sum beamforming (DSB) of photoacoustic data does not incorporate a priori spatial sparsity of the imaging target. By incorporating this information into beamforming for limited-view photoacoustic computed tomography, we experimentally obtained enhanced resolution images of wires at a depth of 8.5 mm in a tissue mimicking scattering medium. Using a 21 MHz transducer, we improved the resolution from the 200 to 250 μm achieved by DSB to 75 μm. The sparsity-based technique also generated a cleaner image with a background signal level of roughly -50 dB, much lower than the roughly -18 dB background signal level of DSB.