Optimal data types in optical tomography

Abstract

This paper addresses the problem of image reconstruction in optical tomography with respect to the measurement types used. We demonstrate the difficulty of the simultaneous reconstruction of absorption and diffusion images, by using both a simple circular case with embedded inhomogeneities, and a complex neonatal head model, and show that improvements are possible by combining suitable measurement types. We analyse the potential ability of the reconstruction program to separate absorption and scattering features by plotting the error norm as a function of absorption and scattering coefficient in the single perturbation case for a number of different measurement types.