The utility of optical tomography as a practical imaging modality has thus far been limited by its intrinsically low spatial resolution and quantitative accuracy. Recently we have argued that a broad range of physiological phenomenon might be accurately studied by adopting this technology to investigate dynamic states. One such phenomenon holding considerable significance is the dynamics of the vasculature, which has been well characterized as being both spatially and temporally heterogeneous. In this report we have modeled such heterogeneity in the limiting case of spatiotemporal coincident behavior involving optical contrast features, in an effort to define the expected limits with which dynamic states can be characterized using two newly described reconstruction methods that evaluate normalized detector data; the normalized difference method (NDM) and the normalized constraint method (NCM). Influencing the design of these studies is the expectation that spatially coincident temporal variations in both the absorption and scattering properties of tissue can occur in vivo. We have also chosen to model DC illumination techniques, in recognition of their favorable performance and cost for practical systems. This choice was made with full knowledge of theoretical findings arguing that separation of the optical absorption and scattering coefficients under these conditions is not possible. Results obtained show that the NDM algorithm provides for good spatial resolution and excellent characterization of the temporal behavior of optical properties but is subject to inter–parameter crosstalk. The NCM algorithm, while also providing excellent characterization of temporal behavior, provides for much improved spatial resolution, as well as for improved separation of absorption and scattering coefficients. A discussion is provided to reconcile these findings with theoretical expectations.
Keywords: Dynamic optical tomography, Physiological dynamics, Physiological oscillations, Inverse problem, Crosstalk and non–uniqueness