Recovery of Raman or Fluorescence signatures from within thin tissues benefits from model-based estimation of where the signal came from, especially if the signal passes through layers in which the absorption or scattering signatures distort the signal. Estimation of the signal strength requires appropriate normalization or model-based recovery, but the key to achieving good results is a good model of light transport. While diffusion models are routinely used for optical tomography of tissue, there's some thought that more precise radiation transport modeling is required for accurate estimation. However, diffusion is often used for small animal imaging, because it's a practical approach, which doesn't require knowledge of the scatter phase function at each point in the tissue. The question asked in this study is, whether experimentally acquired data in small volumes such as a rodent leg can be accurately modeled and reconstructed using diffusion theory. This study uses leg geometries extracted from animal CT scans and liquid phantoms to study the diffusion approximations. The preliminary results show that under certain conditions the collected data follows the expected trend.