Development of two-dimensional beam hardening correction for X-ray micro-CT.

Beam hardening is a well-known issue in X-ray tomography that arises due to the nonlinear interaction of polychromatic X-rays with materials. This effect distorts reconstructed images, especially when multiple materials are present. Prior research has shown that using a step wedge with a single material can linearise attenuation data. However, this method fails when materials with dissimilar X-ray attenuation properties are involved. That uncertainty drove the need for a more generalised correction method. No prior work had resolved the challenge of correcting beam hardening in two-phase biological samples. This gap motivated the development of a new approach that accounts for multiple materials simultaneously. The problem is particularly relevant in biological imaging, where samples are often immersed in preservation fluids. Existing methods require segmentation of each material phase, which is time-consuming and error-prone. This study aimed to address these limitations by proposing a novel correction strategy.

The goal of this research was to improve beam-hardening correction in X-ray micro-CT for two-phase samples. Specifically, the study focused on biological hard tissues that are either immersed or embedded in other media. Traditional linearisation methods are inadequate when materials with distinct attenuation properties are present. The authors aimed to develop a correction method that does not require manual segmentation of each phase. This approach would enhance the accuracy of attenuation measurements in complex samples. The study also sought to test the proposed method on a hydroxyapatite disk immersed in ethanol. The motivation stemmed from the limitations of current techniques in biological imaging. By addressing these issues, the study aimed to provide a more reliable and efficient correction method for X-ray micro-CT.

The study introduced a two-dimensional step wedge approach for beam-hardening correction. This method was not created physically but was derived from published X-ray attenuation coefficients and a modelled X-ray spectrum. The X-ray spectrum was optimised using measurements from a calibration carousel. The two-phase correction was tested on a hydroxyapatite disk scanned in two conditions: dry and immersed in 70% ethanol. The disk was chosen as a model biological tissue with known attenuation properties. The dry scan served as a baseline for comparison. The immersed scan simulated typical conditions in biological specimen preservation. The method was evaluated by comparing the attenuation coefficients from both scans. The results were assessed using the deviation from the dry case as a metric for correction accuracy.

The two-dimensional beam-hardening correction significantly improved accuracy in the immersed hydroxyapatite disk. Simple linearisation resulted in a 10% deviation in edge attenuation between the dry and immersed scans. In contrast, the two-dimensional correction reduced this deviation to only 0.5%. The method effectively accounted for the differences in X-ray attenuation between the disk and the ethanol solution. This outcome demonstrated the method's ability to correct for beam hardening in two-phase samples. The correction did not require individual segmentation of the two phases. The results showed that the method could be applied without prior knowledge of material boundaries. The accuracy of the corrected attenuation coefficients was comparable to the dry case. These findings suggest the method is robust for biological imaging applications.

The authors concluded that the two-dimensional beam-hardening correction method is effective for two-phase samples in X-ray micro-CT. The method does not require segmentation of individual phases, which is a limitation of traditional approaches. The study demonstrated that the method can correct for beam hardening in immersed biological tissues. The accuracy of the correction was validated using a hydroxyapatite disk in ethanol. The results showed a significant reduction in attenuation deviation compared to simple linearisation. The method's success was attributed to the use of a modelled X-ray spectrum and published attenuation coefficients. The approach is particularly useful for biological samples preserved in immersion solutions. The findings suggest that this method can improve the reliability of X-ray micro-CT in such contexts.