Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Retrieval of multiple fibre orientations using X-ray dark-field signal modelling

Lorenzo Massimi, Michela Fratini, Shashidhara Marathe, Christoph Rau, Giuseppe Gigli, Alessandro Olivo, Alessia Cedola

Abstract

Dark-field imaging is widely used to infer fibre orientation from signal modulation as a function of sample orientation. However, current X-ray dark-field retrieval methods are restricted to single orientations and require tomography to resolve overlapping structures. This approach is time-consuming and not suitable for thin materials, which are common in materials science. Here we present a dark-field model capable of retrieving multiple fibre orientations within a single pixel. The model, based on a geometrical description of fibre scattering, was validated through Monte Carlo simulations and experiments using beam-tracking setups with 1D and 2D masks. Results demonstrate reliable orientation retrieval for up to two fibres per pixel, with the 2D mask providing multi-directional sensitivity in a single acquisition and enabling faster and simplified data collection.

Retrieval of multiple fibre orientations using X-ray dark-field signal modelling

Abstract

Dark-field imaging is widely used to infer fibre orientation from signal modulation as a function of sample orientation. However, current X-ray dark-field retrieval methods are restricted to single orientations and require tomography to resolve overlapping structures. This approach is time-consuming and not suitable for thin materials, which are common in materials science. Here we present a dark-field model capable of retrieving multiple fibre orientations within a single pixel. The model, based on a geometrical description of fibre scattering, was validated through Monte Carlo simulations and experiments using beam-tracking setups with 1D and 2D masks. Results demonstrate reliable orientation retrieval for up to two fibres per pixel, with the 2D mask providing multi-directional sensitivity in a single acquisition and enabling faster and simplified data collection.
Paper Structure (3 sections, 3 equations, 4 figures, 2 tables)

This paper contains 3 sections, 3 equations, 4 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Methods
  3. Results

Figures (4)

  • Figure 1: (a, b) Schematic view of BT setup using 1D and 2D absorption mask, respectively.
  • Figure 2: (a) Simulated DF images of a carbon-composite phantom at different rotation angles. (b) Angular intensity profiles from the ROIs in (a) with retrieved fibre orientations overlapped to a crop of simulated image (inset). (c) Retrieved orientations for one, two, and three fibre ROIs at varying noise levels shown as box plots, with fitted profiles in the inset for signal quality comparison.
  • Figure 3: Phase-retrieval procedures for both masks (a,b), and retrieved DF images at different mask–sample angles (c,d) with corresponding intensity profiles from ROIs containing one, two, or three layers.
  • Figure 4: Panels (a,d) show the fit of the mean dark-field intensity (insets) for the 1D and 2D masks, respectively. (b) probability maps for each pixel to belong to background, one, two, or three layer regions. (c,e) binarized maps assigning the most probable layer count to each pixel; retrieved fibre orientations for ROIs 1–3 are shown in the side panels (single fibre in yellow, two fibres in red, three fibres in white). (d,g) Quantitative comparison of fibre orientations retrieved from one, two, and three layer regions for the 1D and 2D masks, respectively.