Quantitative measurement on three-dimensional computed tomography: an experimental validation using phantom objects.

BACKGROUND: The use of 3-dimensional computed tomography (CT) imaging has been applied to the craniofacial region as well as to many other parts of the human body. Quantitative measurements have frequently been performed on the 3-dimensional images. However, critical validation of the measurement has been insufficient in the literature. This study was designed to evaluate the errors of the 3-dimensional measurements. METHODS: Four phantom objects, a cube, a sphere, a cylinder, and a life-size adult skull model, were scanned using standard CT acquisition protocol. The data were transferred, reformatted, and displayed on an IBM-compatible personal computer running AnalyzePC 2.5 software. Linear, area, and volume measurements were obtained using one of the two methods. The first was physical measurement of the phantom objects using a caliper for linear measurement and mathematical calculations for area and volume measurements. The second was done by computer measurement on 3-dimensional images using the AnalyzePC 2.5 program. Each measurement was performed twice. The differences were compared between the repeated measurements and between the two methods. RESULTS: The images were displayed according to standard 3-dimensional CT protocol. The differences between the measurements were insignificant and ranged from 0.00 to 2.57%. CONCLUSION: This study validated the accuracy of the quantitative measurements on 3-dimensional CT images.