Scatter Correction Methods in Dimensional CT

The purpose of this study was to evaluate different approaches for scatter artifact reduction in the field of dimensional computed tomography (CT). The methods we investigated can be subdivided into one software-based method and three hardware-based methods. The software-based method uses a Monte Carlo scatter simulation to estimate scatter intensities for an initial model of the geometrical and physical properties of the object and the CT measuring machine. On the hardware side, we evaluated the influence of an anti-scatter grid with variable grid ratios, the impact of different beam collimations, and a water bath with the object immersed inside. Our preliminary results show that all tested methods are more or less successful in reducing scatter artifacts that show up in reconstructed CT images. Among all methods the water bath approach showed the worst performance in terms of scatter reduction. This is accompanied by the disadvantage that the object must be immersed in a material of similar density, which further devalues the practical use of this method. The slit scan method showed good results for scatter artifact reduction but has the disadvantage that multiple scans must be performed to measure the full 3D volume. Thus far, the Monte Carlo based method and the use of anti-scatter grids offer the best results for scatter reduction and practical application. They reduce scatter artifacts to a great extent and offer vastly improved images. Furthermore, these two methods do not complicate the process of data acquisition as the water bath and the slit scan method do.