Frequency domain method for photoacoustic tomography with an arbitrary detection geometry

Photoacoustic imaging is a high resolution, biomedical imaging modality used for cancer detection and imaging biological tissues. Exact photoacoustic reconstruction methods were proposed for simple geometries in both time and frequency domain. However, for arbitrary geometries only time reversal methods were previously proposed. In this paper, we propose a frequency domain algorithm for photoacoustic image reconstruction given an arbitrary sensor geometry. The spatial source distribution is expanded using a 2D Fourier Bessel expansion and the proposed method estimates Fourier Bessel coefficients of this expansion using a least squares method at frequencies corresponding to the zeros of Bessel functions. The effectiveness of the proposed method was corroborated by numerical experiments.