Numerical investigation of nonlinear sound propagation of photoacoustic tomography imaging.

Almost all reconstruction methods in photoacoustic tomography (PAT) have been developed by assuming that sound propagation is linear, which is valid for ordinary PAT applications but would become inappropriate when the sound amplitude is higher than a certain threshold level. In the current study, we investigate the effect of nonlinear sound propagation on PAT by using a numerical method which utilizes the time-reversal (TR) technique. In the forward stage, the Euler equations are solved to simulate nonlinear sound propagation, and the flow variables (pressure, velocity and density) are recorded by an array of virtual sensors. The recorded data are used to reconstruct the initial fields within the computational domain by using both linear and nonlinear TR techniques. Furthermore, TR results constructed with and without the recorded flow velocity field, which is difficult to measure for practical applications, have also been compared. The current results show that nonlinear reconstructions produce images with superior clarity, resolution and contrast compared to those reconstructed by the linear method, particularly when the recorded velocity field is used in the reconstruction.