Multiview linear-array photoacoustic computed tomography with improved elevational resolution in 3D

Linear transducer arrays are widely used in photoacoustic computed tomography (PACT) due to their lightweight design, cost-effectiveness, and handheld convenience. Despite these advantages, the inherent focusing design of the linear array results in anisotropic resolution in 3D. Specifically, the elevational resolution is approximately an order of magnitude worse than both lateral and axial resolutions. In this work, we propose multiview linear PACT (MvLPACT) to overcome this limitation. By combining the scanned data along the elevational direction from multiple angles, MvLPACT compensates for the deficiencies in elevational resolution. A multiview fast iterative shrinkage-thresholding algorithm (FISTA) based deconvolution, named MvFISTA-Deconv, is applied to the multiview scanning data to reconstruct 3D images. MvFISTA-Deconv integrates the high-frequency components that correspond to high-resolution features acquired from multiple angles into the reconstructed 3D images with improved resolution. Our proposed MvLPACT is evaluated on both phantoms and in vivo human experiments, showing a significant elevational resolution improvement up to 8-fold. Our results demonstrate MvLPACT as a powerful and robust technique for high-resolution 3D imaging in PACT applications, offering substantial advancements over traditional linear-array-based PACT.