Compressive sensing photoacoustic imaging based on multi-view error gradient fusion

Compressive sensing photoacoustic imaging (PAI) is an emerging method for photoacoustic signal sampling and image reconstruction. Based on the method, the high-imaging rate and low-system cost can be achieved. Unfortunately, artefacts exist in the reconstructed images and appear heavier if the measurement is fewer. The artefacts appear at different pixels when images are observed from different angles, and can be rejected by the fusion of multi-view images. In this study, a multi-view image fusion method based on error gradient is proposed. The theory of compressive sensing PAI is introduced firstly. Then a feasible measurement scheme based on the digital micromirror device is constructed. Subsequently, error gradient of the reconstructed image is defined and the fusion rule based on error gradient is made. Finally, the proposed method is testified by the simulations, in which images observed from 0 and 90° are used to take the fusion. Simulation results demonstrate that the image artefacts can be reduced effectively by the proposed method.