Correlation between 3D imaging methods in studying bone architecture: SEM, microCT and confocal LM

abstract

English

Computerised x-ray microtomography (microCT) is increasingly used in the 3D study of bone microarchitecture and in quantifying bone volume fractions. However, the volumetric resolution in labora- tory apparatus for small rodent studies is at best several microns linear and recognisable detail characterising forming, resting and resorbing surfaces is completely missing. Backscattered electron mode scanning electron microscopy (BSE SEM) of both macerated 3D samples and polished surfaces of blocks of PMMA embedded tissue provides this information but samples have to be cut and processed. The same PMMA material is good for confocal fluores- cence microscopy (CSLM) for both tissue morphology and the study of tetracycline and calcein mineralising front labels to 50– 200 microns deep to the block surface. With the recent acquisi- tion of SEM with variable chamber pressure to permit examina- tion of uncoated specimens, we are able to conduct CSLM after SEM for correlation studies. Here, we report new approaches to correlation between all these imaging methods in the study of 70% ethanol fixed normal femurs from 330 g male Lewis rats from another study. Microtomography used a Scanco microCT 40 sys- tem using 45 or 55 kV and 8 lm linear voxel size. The distal femur was bisected longitudinally with a water cooled diamond saw, one half was macerated in alkaline pronase and the other embed- ded in PMMA prior to 20 kV 3D BSE imaging. Embedded block faces were imaged uncoated in the SEM and by confocal micros- copy. Volumetric data analysis used ImageJ. Drishti software (Aus- tralian National University) was used to reconstruct views corresponding to the SEM images with particularly good matches between SEM and microCT for the macerated trabecular bone. Exten- sive regions of thin trabeculae were frequently missing in the microCT reconstruction. This is clearly a partial volume problem but it draws attention to the fact that an extensive network of fine tra- beculae are lost to microCT visualisation and analysis on a routine basis. These fine rods are frequently 7 and down to < 2 microm in gauge, well below the detection limit for microCT. Their existence overturns results and theories about interconnectedness.

x-ray microtomography,scanning electron microscopy,confocal fluorescence microscopy,rat femur

ZIKMUND, T.; KVASNICA, L.; BOYDE, A.

1. 7. 2012

0021-8782

JOURNAL OF ANATOMY

221

1

United Kingdom of Great Britain and Northern Ireland

86

86

1