Abstract
Abstract Image reconstruction for photoacoustic tomography is affected by the finite aperture effect. The finite-aperture effect is caused by spatial impulse response associated with finite-sized acoustic transducer, resulting in the worse tangential resolution when imaging points are close to the transducer position. In this study, we developed a model-based image reconstruction technique for photoacoustic tomography to solve such a problem. Based on a linear, discrete PAT imaging model, the proposed method employs a spatiotemporal optimal filter designed in minimum mean square error sense to compensate the SIRs associated with an unfocused and focused transducer at every imaging point; thus retrospective restoration of the tangential resolution can be achieved. Simulation and experimental results demonstrate that this method can substantially improve the degraded tangential resolution for PAT with finite-sized unfocused transducers while retaining the radial resolution. In actually, doing model-based image reconstruction contains many limitations leading to low efficiency, because it requires a lot amount of memories. For cooperating to deal with experimental data, we propose a method to simplify complexity of the algorithm of model-based image reconstruction. It would be adjusted from the condition of experiment and increases the range of observation. In future, We hope to further improve model-based image reconstruction to reduce the computational complexity. It increases the algorithm's usefulness.