Single-View Tomographic Reconstruction Using Learned Primal Dual

TL;DR

This paper evaluates the Learned Primal Dual method for single-view tomographic reconstruction of axially-symmetric objects, demonstrating its effectiveness under challenging conditions with limited data.

Contribution

It extends the application of LPD to extreme single-view scenarios and compares its performance against traditional methods in two X-ray modalities.

Findings

LPD outperforms traditional inversion methods in single-view reconstructions.

LPD maintains robustness under noise and blur conditions.

Effective in both parallel and cone-beam X-ray modalities.

Abstract

The Learned Primal Dual (LPD) method has shown promising results in various tomographic reconstruction modalities, particularly under challenging acquisition restrictions such as limited viewing angles or a limited number of views. We investigate the performance of LPD in a more extreme case: single-view tomographic reconstructions of axially-symmetric targets. This study considers two modalities: the first assumes low-divergence or parallel X-rays. The second models a cone-beam X-ray imaging testbed. For both modalities, training data is generated using closed-form integral transforms, or physics-based ray-tracing software, then corrupted with blur and noise. Our results are then compared against common numerical inversion methodologies.