Quantitative dual-energy micro-CT with a photon-counting detector for material science and non-destructive testing

Abstract

The recent progress in photon-counting detector technology using high-Z semiconductor sensors provides new possibilities for spectral x-ray imaging. The benefits of the approach to extract spectral information directly from measurements in the projection domain are very advantageous for material science studies with x-rays as polychromatic artifacts like beam-hardening are handled properly. Since related methods require accurate knowledge of all energy-dependent system parameters, we utilize an adapted semi-empirical model, which relies on a simple calibration procedure. The method enables a projection-based decomposition of photon-counting raw-data into basis material projections. The objective of this paper is to investigate the method's performance applied to x-ray micro-CT with special focus on applications in material science and non-destructive testing. Projection-based dual-energy micro-CT is shown to be of good quantitative accuracy regarding material properties such as electron densities and effective atomic numbers. Furthermore, we show that the proposed approach strongly reduces beam-hardening artifacts and improves image contrast at constant measurement time.

Fig 1. Experimental setup.

Photography of the spectral micro-CT set-up: (A) micro-focus x-ray source with beam collimator, (B) linear-stages with calibration phantoms to automatically calibrate the model parameters, (C) mounting and positioning devices for the sample and adjustment of magnification and voxel-size, (D) photon-counting detector.

Fig 2. Decomposition accuracy.

In order to asses the reliability of the calibration of the forward-model in Eq 4 a testgrid different from the calibration grid was acquired (A). The measured dual-energy data was decomposed into basis material thicknesses (B) and compared to the ground truth (GT) values for Ti (C) and POM (D), respectively.

Fig 3. Spectral CT images of the phantom used to assess quantitative accuracy.

While the polychromatic attenuation image (A) shows typical beam-hardening artifacts, the basis material images (B,C) are unaffected of those and yield a clear decomposition into volume fractions of Ti and POM. The VMI at 80 keV (D) alongside with the effective atomic number (E) and electron density image (F) provide quantitative information of the studied object. The accuracy of the extracted quantities is evaluated in Table 2 and Fig 4. The depicted images correspond to the mean of 10 slices of the reconstructed image volume to improve the visual appearance of potential artifacts.

Fig 4. Quantitative accuray of virtual monoenergetic images.

The VMIs calculated from the obtained basis material images (c. f. Fig 3B and 3C)) are compared to the theoretical attenuation values (label GT) of the studied materials (A) in the energy range from 40 to 200 keV. (B) shows the relative deviation between the values extracted from the measurement and the corresponding ground truth values. The theory values were taken from the XCOM database [31].

Fig 5. Spectral CT images of the phantom used to assess the capability of correcting beam-hardening.

The upper row (A) shows the reconstructed image data after a copper wire was added to the phantom. The influence of the strongly attenuating copper wire was analyzed by line-plots (B). The depicted images correspond to the mean of 10 slices of the reconstructed image volume to improve the visual appearance of potential artifacts.

Fig 6. Conventional and spectral CT images of a concrete drill core.

The conventional image (A) depicts the attenuation of the polychrmoatic spectrum, while the decomposed basis material images show the volume fractions of titanium (B) and POM (C). (D) gives the sample’s attenuation value at 70 keV. (E) and (F) depict the spatial distribution of the effective atomic number and the electron density within the sample, respectively.

Fig 7. Conventional and spectral CT images of an ethernet connector.

The top row depicts polychromatic attenuation images, while the bottom row shows VMIs of the corresponding sample regions. Subfigure (B) and (E) show more narrow windowed clippings of (A) and (D), respectively. (C) and (F) depict a different position inside the reconstructed volume. The energy levels of the shown VMIs were chosen in favor of best visual appearance.