MicroCT Facility (MCT) X-Ray Computed Microtomography X-Ray Micro-Computed Tomography

Sample Diameter

The scanner can physically accommodate widths of 500 mm, but the field of view (FOV) of the detector cannot capture that entire width. Therefore, scanning a large sample will require region of interest (ROI) scanning, or scanning only a portion of the sample rather than the whole sample volume. ROI scanning is an option, but the user should be aware that this option can cause artifacts near the edges of the image especially for non-homogenous, non-cylindrical samples. Usually, it is best to keep the entire width of the sample within the FOV of the detector at all angles of rotation.

For non-ROI scans on the currently installed detector options, the maximum practical diameter of a sample is about 110 mm. This diameter can essentially be doubled (to 220 mm) if tiling or offset scans are used. Keep in mind that file size and acquisition time will also be doubled when using tiling or offset scans.

The graph below shows the relationship between the FOV width (which can be thought of as the width of the sample) and the maximum theoretical voxel size (VS) of the resulting image. The FOV width can be doubled if tiling or offset scanning is used. The main takeaway of the graph is that, without ROI scanning, there is a maximum achievable voxel size for a given sample diameter.

Graph showing the relationship between voxel size and field of view for the UniTOM HR

The user should also be aware that voxel size does not equate to spatial resolution. In an ideal case voxel size and spatial resolution would be equivalent, but many factors degrade this equivalency including x-ray spot size and drift, sample movement, image contrast, signal-to-noise ratio, image artifacts and partial volume effects. As a rule of thumb, to qualitatively identify features in a microCT image, the feature diameter (D) needs to be at least two to five times the voxel size. But, to quantify features they need to be at least five to twenty times larger than the voxel size. The graph below displays this rule of thumb.

Graph showing the relationship between feature diameter and voxel size for the UniTOM HR

Using these guidelines, a user can estimate the smallest feature that will be quantifiable for a given sample width. If that resolution is unacceptable for their research goals, the user will need to decrease their sample diameter or utilize ROI scanning and be comfortable with the possibility of image artifacts.

Another factor to consider is the sample composition. A highly dense sample may require a smaller diameter to allow the collection of a microCT image in a reasonable amount of time.