Sample Composition: Similar-Z Materials

Similar-Z Materials in the Same Sample

Another difficult sample is one in which the phases in a sample are of similar x-ray attenuations. This is often an issue in low-Z materials such as organics where many of the organic structures in a sample have similar elemental compositions making it hard to differentiate them in a tomographic image.

• Limited Contrast
  • The main issue with this type of sample is that contrast between the phases is minimal or negligible. This causes issues during image processing where many algorithms will have a hard time separating the phases for 3D measurements.

Possible Solutions

1. A possible way to mitigate this issue of limited contrast between phases is by using phase contrast imaging. This type of imaging utilizes the refractive index of the different phases to effectively thicken the phase boundaries in the resulting CT image. This can only be accomplished in microCT systems that can accommodate a large distance (on the order of at least ~1 m) between the sample and the detector unless the system has a built in modulator. This is because the X-rays are reflected at a small angle and a large distance is needed for the reflected X-rays to make any measurable difference in the resulting image. A detector with smaller pixel pitch can also decrease the sample-detector distance needed to observe phase contrast. 
2. Another option is to use a contrast agent. This is similar to optical microscopy staining, where the stain attaches to the structure of interest allowing for a high contrast image, but a microCT contrast agent effectively changes the attenuation of a phase in the sample to increase the contrast in the resulting CT image. One common example is the use of a contrast agent in experiments involving the unsaturated flow of water through porous media. In this example, a salt such as potassium iodide is added to the water to enhance the contrast between the water and other fluids in the porous medium. This option isn't always ideal, though, because it can be difficult to make the contrast agent prioritize a specific phase in many systems. Also it inherently changes the composition of the sample, which isn't always an option if the sample needs to be preserved.
3. Sometimes machine learning segmentation can be used to separate phases during processing of low contrast images. Advances in machine learning techniques have allowed many researchers to use data that would normally have been very difficult to segment by conventional techniques. It should be noted that this technique only works if there is at least some contrast between the phases, so it is often used in conjunction with phase contrast imaging.