A Bayesian approach for high resolution imaging of small changes in multiple scattering media

This paper introduces a Bayesian approach to achieve high-resolution imaging of sub-wavelength changes in the presence of multiple scattering. The approach is based on the minimization of a cost function defined by the decorrelations induced in the measured waveforms by the apparition of a local changes. Minimization is achieved via a Monte Carlo Markov Chain (MCMC) algorithm combined to an analytical model that computes the sensitivity kernel of the medium. In the inversion procedure, the parameters to infer represent the physics of the problem, such as the diffusivity in the medium and/or the geometrical features of the reflector (position and scattering cross-section). The method is successfully compared to the linear inversion approach initially proposed for the so-called Locadiff imaging method through several examples, both numerical and experimental.