Biological tissue is a highly scattering medium that prevents deep imaging of light. For medical applications, optical imaging offers a molecular sensitivity that would be beneficial for diagnosing and monitoring of diseases. Acousto-optical tomography has the molecular sensitivity of optical imaging with the resolution of ultrasound and has the potential for deep tissue imaging. Here, we present a theoretical study of a system that combines acousto-optical tomography and slow light spectral filters created using spectral hole burning methods. Using Monte Carlo simulations, a model to obtain the contrast-to-noise ratio (CNR) deep in biological tissue was developed. The simulations show a CNR > 1 for imaging depths of ∼5 cm in a reflection mode setup, as well as, imaging through ∼12 cm in transmission mode setups. These results are promising and form the basis for future experimental studies.