k−space models for broadband acoustic pulse propagation differ from pseudo−spectral time domain (PSTD) models in their treatment of the time step. By replacing a finite−difference scheme with a propagator, exact for homogeneous media, larger time steps can be taken without loss of accuracy or stability and without introducing dispersion. Three k−space models for modeling photoacoustically generated (PA) pulses are described here. A very simple, exact, model of PA propagation in a homogeneous fluid is used to introduce the k−space propagator, and two models of propagation in heterogeneous media, originally designed for modeling scattering in soft tissue, are adapted for use in photoacoustics [Mast et al., IEEE Trans. UFFC 48, 341−354 (2001); Tabei et al., J. Acoust. Soc. Am. 111, 53−63 (2002)]. Our motivation for describing these models comes from biomedical PA imaging, in which one of the current limitations is the assumption that soft tissue has a uniform sound speed. Efficient, accurate, and simple−to−encode forward models such as these are very useful for studying the effects of the heterogeneities encountered in practice. They may also be useful in designing PA imaging schemes that can account for acoustic heterogeneities. [This work was funded by the EPSRC, UK]
Read full abstract