Tensor-Newton Reconstruction Scheme for Fluorescence Optical Tomography

In this work, we set up the use of a Tensor-Newton scheme for solving the nonlinear reconstruction problem for the fluorophore absorption coefficient, μaf
^x
in SPN modeled fluorescence optical tomography. We present numerical reconstruction studies for differential uptake of fluorophore with noiseless and noisy data in test cases with varying contrast between the object and background μaf
^x
value. Comparisons are presented between a first order regularising Levenberg-Marquardt scheme and two variants of the Tensor-Newton scheme. The Tensor Newton scheme is found to be more robust and performs mostly better than and at at least at par with the first order scheme in the test cases considered.