Aberration correction in photoacoustic imaging using paraxial backpropagation

Photoacoustic (PA) imaging aims for the reconstruction of acoustic sources that originate in pulsed light absorption. While the speed of sound (SOS) in biological tissue is heterogeneous, standard reconstruction algorithms usually assume a constant SOS. This results in deformations of the reconstructed sources, which are referred to as aberration and can have a severe impact on the spatial resolution. If the SOS is known, aberrations can be compensated during the reconstruction at the expense of an increasing computational effort. Algorithms that account for heterogeneous SOS usually compute individual delays for each reconstructed pixel or perform an entire wave field simulation. In this contribution, we present an alternative approach by introducing a PA backpropagation based on a paraxial approximation of the wave equation that can be computed with almost the computational efficiency of a standard frequency domain reconstruction (FSAFT). The method accounts for refraction and neglects unwanted back-reflections during the backpropagation process.