DEPICT | Material Decomposition in X-ray Phase-Contrast Imaging with Coherent Sources

Summary
X-ray microtomography is increasingly popular at synchrotron sources worldwide. The high flux and excellent beam properties allow for a fast three-dimensional visualization of various kinds of specimens at micrometre spatial resolution and exceptionally high image quality. But all these imaging instruments use the unique synchrotron sources merely to depict the objects’ internal structure. All further quantitative information encoded into the data remains unused. The reason is a lack of an image retrieval method, which accurately models the whole image formation process with highly-coherent radiation on the micrometre scale to extract all contained information from a single exposure. Consequently, these microtomographic images always contain a mixture of different signals based on various interaction effects – namely attenuation, refraction and scattering. Accurately separating all the signals will allow to quantify concentration and microstructure of materials of interest and to locate them inside the object, which is unattainable in a single exposure at the moment.

The main goal of this project is to enable simple and fast X-ray microtomography at highly-coherent sources to fully exploit the disregarded information contained in the imaging data. My team will develop a robust algorithm for accurate signal separation modelling all interaction effects and demonstrate the benefit of material decomposition for highlighted applications. With this tool at hand our ambition is to fundamentally change the way of imaging at modern synchrotron sources: Imaging will no longer be solely a morphological method, as quantitative and material-specific information will become standardly available in fast microtomography at all coherent sources worldwide.
Results, demos, etc. Show all and search (0)
Unfold all
/
Fold all
More information & hyperlinks
Web resources: https://cordis.europa.eu/project/id/101125761
Start date: 01-03-2024
End date: 28-02-2029
Total budget - Public funding: 2 620 750,00 Euro - 2 620 750,00 Euro
Cordis data

Original description

X-ray microtomography is increasingly popular at synchrotron sources worldwide. The high flux and excellent beam properties allow for a fast three-dimensional visualization of various kinds of specimens at micrometre spatial resolution and exceptionally high image quality. But all these imaging instruments use the unique synchrotron sources merely to depict the objects’ internal structure. All further quantitative information encoded into the data remains unused. The reason is a lack of an image retrieval method, which accurately models the whole image formation process with highly-coherent radiation on the micrometre scale to extract all contained information from a single exposure. Consequently, these microtomographic images always contain a mixture of different signals based on various interaction effects – namely attenuation, refraction and scattering. Accurately separating all the signals will allow to quantify concentration and microstructure of materials of interest and to locate them inside the object, which is unattainable in a single exposure at the moment.

The main goal of this project is to enable simple and fast X-ray microtomography at highly-coherent sources to fully exploit the disregarded information contained in the imaging data. My team will develop a robust algorithm for accurate signal separation modelling all interaction effects and demonstrate the benefit of material decomposition for highlighted applications. With this tool at hand our ambition is to fundamentally change the way of imaging at modern synchrotron sources: Imaging will no longer be solely a morphological method, as quantitative and material-specific information will become standardly available in fast microtomography at all coherent sources worldwide.

Status

SIGNED

Call topic

ERC-2023-COG

Update Date

12-03-2024
Images
No images available.
Geographical location(s)