Consistent Inversion of Noisy Non-Abelian X-Ray Transforms

For a simple surface, the non-linear statistical inverse problem of recovering a matrix field from discrete, noisy measurements of the -valued scattering data of a solution of a matrix ODE is considered (). Injectivity of the map was established by [Paternain, Salo, Uhlmann; Geom.Funct.Anal. 2012]. A statistical algorithm for the solution of this inverse problem based on Gaussian process priors is proposed, and it is shown how it can be implemented by infinite-dimensional MCMC methods. It is further shown that as the number of measurements of point-evaluations of increases, the statistical error in the recovery of converges to zero in -distance at a rate that is algebraic in , and approaches for smooth matrix fields . The proof relies, among other things, on a new stability estimate for the inverse map . Key applications of our results are discussed in the case to polarimetric neutron tomography, see [Desai et al., Nature Sc.Rep. 2018] and [Hilger et al., Nature Comm. 2018]
View on arXiv