Boundary effects on image reconstruction in photoacoustic tomography

Exact photoacoustic tomography requires scanning over a 4&pgr; solid angle in 3D. The ultrasound detection window, however, is often limited, which makes a full scan impossible. For example, when a boundary lies closely to an object, the scanning region can cover only less than 4&pgr; in 3D. Because of incomplete information, the resolution, SNR, and fidelity of the resulting image deteriorate. Boundaries, however, can be used to our advantage; we proposed post-processing algorithms in image reconstruction to make partially scanned data complete. Here, we show the efficacy of the post-processing algorithms with both numerical and experimental results. Indeed, the algorithms can improve the resolution, SNR, and fidelity.