English

Efficient Nonparametric Bayesian Inference For X-Ray Transforms

Statistics Theory 2018-02-14 v2 Analysis of PDEs Methodology Statistics Theory

Abstract

We consider the statistical inverse problem of recovering a function f:MRf: M \to \mathbb R, where MM is a smooth compact Riemannian manifold with boundary, from measurements of general XX-ray transforms Ia(f)I_a(f) of ff, corrupted by additive Gaussian noise. For MM equal to the unit disk with `flat' geometry and a=0a=0 this reduces to the standard Radon transform, but our general setting allows for anisotropic media MM and can further model local `attenuation' effects -- both highly relevant in practical imaging problems such as SPECT tomography. We propose a nonparametric Bayesian inference approach based on standard Gaussian process priors for ff. The posterior reconstruction of ff corresponds to a Tikhonov regulariser with a reproducing kernel Hilbert space norm penalty that does not require the calculation of the singular value decomposition of the forward operator IaI_a. We prove Bernstein-von Mises theorems that entail that posterior-based inferences such as credible sets are valid and optimal from a frequentist point of view for a large family of semi-parametric aspects of ff. In particular we derive the asymptotic distribution of smooth linear functionals of the Tikhonov regulariser, which is shown to attain the semi-parametric Cram\'er-Rao information bound. The proofs rely on an invertibility result for the `Fisher information' operator IaIaI_a^*I_a between suitable function spaces, a result of independent interest that relies on techniques from microlocal analysis. We illustrate the performance of the proposed method via simulations in various settings.

Keywords

Cite

@article{arxiv.1708.06332,
  title  = {Efficient Nonparametric Bayesian Inference For X-Ray Transforms},
  author = {François Monard and Richard Nickl and Gabriel P. Paternain},
  journal= {arXiv preprint arXiv:1708.06332},
  year   = {2018}
}

Comments

38 pages, 6 figures

R2 v1 2026-06-22T21:19:49.409Z