Managing Errors with X-Ray Computed Tomography (X-Ray CT) When Measuring Physical Properties

ABSTRACT X-ray computer tomography (CT) measures X-ray absorption statistics (mean and standard deviation) for selectable regions of solid objects. These statistics are useful because they frequently relate to physical properties such as moisture and density. Scans are expensive and it is therefore desirable to conduct studies with minimum scans for a set level of precision. A rational decision aid for determining the number of scans was presented. A rationale was also presented for removing variation due to noise from the standard deviation statistic. Using correlation analyses, subregion radius should exceed 4-8 pixels to maintain independence in observations. Although these approaches embody some assumptions as yet unverified, they provide guidance in planning and interpreting CT based experiments.