Comparison of microcomputed tomographic and microradiographic measurements of cortical bone porosity

• Published in: Calcified tissue international 2004-05, Vol.74 (5), p.437-447
• Main Authors: Cooper, D M L, Matyas, J R, Katzenberg, M A, Hallgrimsson, B
• Format: Article
• Language: English
• Subjects: Femur - diagnostic imaging; Femur - ultrastructure; Humans; Image Processing, Computer-Assisted; Microradiography - methods; Reproducibility of Results; Tomography, X-Ray Computed
• ISSN: 0171-967X; 1432-0827
• DOI: 10.1007/s00223-003-0071-z

Cortical bone is perforated by a network of canals that have a significant impact upon its material properties. Microcomputed tomography offers the possibility of noninvasively visualizing and quantifying cortical pores in both two and three dimensions. Establishing how two-dimensional (2D) microcomputed tomographic (microCT) analysis compares with conventional methods for analyzing cortical porosity is an important prerequisite for the wider adoption of this technique and the development of three-dimensional (3D) analysis. Therefore, we compared porosity-related parameters from 2D microcomputed tomographic images with those from matching microradiographic sections. Samples from five human femora were scanned at a 10-microm resolution and then sequentially sectioned and microradiographed. An average of eight image pairs were produced from each femur (total, n = 41). The repeatability and comparability of the two techniques was assessed for three parameters; cortical porosity (%), mean pore area (microm(2)), and pore density (pores/mm(2)). For repeatability, no significant difference ( P > 0.05) was found between the two methods for cortical porosity and mean pore area; however, pore density differed significantly ( P < 0.001). For comparability, the bias (+/- error) between the methods was found to be 0.51% (+/-0.31%) for cortical porosity and -155 microm(2) (+/-293 microm(2)) for mean pore area. The bias for pore density was dependent upon measurement size with microcomputed tomographic images having 14% (+/-9.3%) fewer pores per millimeter squared. The qualitative and quantitative similarities between the two techniques demonstrated the utility of 2D microcomputed tomographic for cortical porosity analysis. However, the relatively poor results for pore density revealed that a higher resolution (