IAR - Artifact Reductions in Computed Tomography

Iterative Artifact Reduction IAR

The Iterative Artifact Reduction IAR is a method for the correction of homogeneous components using a characteristic line, which is calculated from the specimen itself by methods of image processing. The patented method is characterized by a number of special features and replaces the extensive characteristic line generation by means of a step wedge.

The calculation of the characteristic line for beam hardening correction is achieved without references. No knowledge about the applied X-Ray spectrum or material properties of the specimen is required. Furthermore IAR considers effects that are not recognized by using a step wedge.

Expansion modules of IAR

Scattered radiation reduction IAR-S is a method that computes the scattered radiation of the specimen analytically. The subtraction of this portion from the original image/projection causes a reduction of the resulting scattered radiation artifacts and consequently an increase of measurement accuracy.

The module IAR-A facilitates an application of the reduction method synchronous to image acquisition. For this purpose previous knowledge is used in terms of a nominal model, so that a characteristic reduction line can already be generated during the acquisition process. This causes an essential acceleration of the characteristic line generation.

Figure 1: The Illustrations show an uncorrected (left) and corrected (right) image of an aluminium specimen with embedded steel pins

With IAR-M a multi material correction is available that is able to reduce artifacts in CT images of heterogeneous components such as e.g. metal or plastics. An improvement concerning the individual material differentiation can be achieved with regard to the surface generation of multi material components. Particularly transitions between individual materials are defined more clearly.

The distributions shown in figure 2 demonstrate that more constant gray scale values are obtained from the objects aluminum portion after the correction. The same applies to the steel portion, whose gray scale values were also homogenized by the reduction.

Figure 2: The graphics denote the gray scale values of the upper profile curves.

Enhancement of measurement accuracy

An artifact reduction causes an increase of measurement accuracy of up to 35% in comparison to the evaluation of uncorrected datasets. In some cases an accurate metrological evaluation is not possible at all without artifact reduction. Figure 3 shows the reduction of the aperture deviation of boreholes inside of an aluminum casting as an example.

Figure 3:reduction of the aperture deviation (left) using boreholes as an example. Schematic demonstration of the large and small boreholes in the aluminium cast part (right)

Industrial computed tomography is, as a non-destructive method, especially appropriate for investigating complex inner and outer component structures with regard to dimensional accuracy and in order to check the production and fabrication tolerances.

System-dependent perturbations and various physical effects reduce the image quality of CT reconstructions. Apart from scattered radiation mainly beam hardening is responsible for that. The reason for this is the polychromatic nature of the X-Ray radiation applied.

The artifacts mentioned express themselves by blurring, loss of contrast and form deviations, which impair image quality and exacerbate or impede the aforementioned testing tasks.