Information transport and limits of optical imaging in the highly diffusive regime

Abstract

Imaging in the highly diffusive regime lies at the heart of various optical medical imaging technologies and other applications such as imaging through fog. However, due to the randomization of the photon propagation direction inside random, complex media, the spatial information of physiological features is degraded such that most diffuse optical imaging techniques are restricted to <10 transport mean free paths (ℓ^{*}). We present an information theoretical analysis of the limits of gathering information in the highly diffusive regime. Our results show that there is still information even when detection is performed at the single-photon level and beyond 200ℓ^{*} (e.g., ∼12 cm tissue). Image reconstructions are enhanced when resolving measurements in both space and time domains. These results provide a general framework for extremely deep diffuse imaging scenarios, such as imaging through the body, and a general context for developing optimized image retrieval strategies.