Variational Photoacoustic Image Reconstruction with Spatially Resolved Projection Data

In this work, we explore different reconstruction algorithms for photoacoustic image reconstruction of spatially resolved projection data. While the commonly used back-projected reconstruction is efficient and fast to compute, it cannot deal with noise that arises during measurements. Therefore, we formulate photoacoustic image reconstruction in a variational framework where we add prior knowledge in terms of Total Generalized Variation. Using this prior knowledge, we can reduce measurement noise and improve the visibility of vessel structures.