X-ray tomography of extended objects: a comparison of data acquisition approaches

TL;DR

This paper compares object stitching and projection stitching methods in X-ray tomography for imaging large objects, focusing on their data acquisition strategies, image quality, and radiation dose efficiency.

Contribution

It provides a comparative analysis of two data acquisition approaches in X-ray tomography, highlighting their advantages, limitations, and practical considerations.

Findings

Projection stitching is more dose-efficient.

Object stitching has easier data alignment.

Projection stitching avoids certain artifacts.

Abstract

The penetration power of x-rays allows one to image large objects. For example, centimeter-sized specimens can be imaged with micron-level resolution using synchrotron sources. In this case, however, the limited beam diameter and detector size preclude the acquisition of the full sample in a single take, necessitating strategies for combining data from multiple regions. Object stitching involves the combination of local tomography data from overlapping regions, while projection stitching involves the collection of projections at multiple offset positions from the rotation axis followed by data merging and reconstruction. We compare these two approaches in terms of radiation dose applied to the specimen, and reconstructed image quality. Object stitching involves an easier data alignment problem, and immediate viewing of subregions before the entire dataset has been acquired. Projection stitching is more dose-efficient, and avoids certain artifacts of local tomography; however, it also involves a more difficult data assembly and alignment procedure, in that it is more sensitive to accumulative registration error.