Figure 1.
Diagram of the imaging process.
A: specimen (photograph and diagram) and specimen stage. Specimens are 5×5×10 mm and the imaged field of view is 2.8×2.8×2.2 mm. B: 2000 radiographs of the specimen are taken from evenly distributed angles from 0° to 180°. C : after tomographic reconstruction, the 3D image is obtained, composed of a stack of slices. D : 256×256 ROI of a reconstructed grey-level image(top), customized non-linear filtering (middle), and segmentation (bottom, microcracks in color, lacunae in white). E: 3D renderings can then be produced to observe and measure microarchitecture or microcracks.
Figure 2.
Illustration of various types of microcracks from 2D slices of SR micro-CT images.
A : long linear microcrack, parallel to the trabecular surface(L1). B: microcrack dividing a trabecula(L2). C : microcrack deflected (black arrow) by the interface between two areas of different mineralization (dots) (L3). D : discrete cracking or bridging. E : tortuous microcrack crossing a lacuna. F : crack (X) perpendicular to the trabecular surface and appearing to be split (black arrow) and crossing a crack (white arrow) driven by a cement line (black dots). G : parallel microcracks contained in a uniformly mineralized area (P). H–I : cross-hatch cracks (x–y plane and x–z plane views) which appear to be parallel in the x–z plane (CH).
Figure 3.
Identification of the same region of the specimen in both SR micro-CT images and epifluorescence microscopy images at low magnification.
A : Original SR micro-CT slice. B : Image computed by averaging the grey levels along the z-axis of the stack. C: Epifluorescence microscopy image showing the same region (green circle) but with a different orientation (cf. red arrows). The black circles contain a microcrack and a pore
Figure 4.
Identification of the same microcracks in images from epifluorescence microscopy at high magnification and SR-micro-CT.
A–B: image from epifluorescence microscopy showing microcracks (white arrows) and an artifact (black arrow). C–D: Corresponding image by SR micro-CT. The orientation of the epifluorescence microscopy images had to be corrected.
Figure 5.
Evaluation of the segmentation method.
A–B: Original SR micro-CT image showing microcracks. Their extremities have been marked with a yellow or red dot. C–D: Segmented images with the extremities of the microcracks reported. The microcracks are well detected and segmented.
Figure 6.
3D renderings of microcracks (in blue) in trabecular bone specimens 4 and 2.
The bone surface is shown in transparent white, revealing the microarchitecture (view from the top the specimen).
Figure 7.
3D renderings of various types of microdamage.
A–B: large and thin linear microcrack (L1), parallel to the surface (front and side view, 512×512×512 ROI). C–D : crack (L2) perpendicular to the surface and matching the shape of the trabecula (front and side view, (512×512×256 ROI). E–F : Parallel cracks(P) (256×256×128 ROI).
Figure 8.
3D renderings of crossed cracks (X), and cross-hatched microcracks (CH).
(512×256×256 ROIs). A: global view of X showing it is actually composed of two twisted cracks, B : zoom (from different angle), C: global view of CH showing it spreads all over the trabecula, D: zoom showing it is made of two series of parallel linear microcracks.
Figure 9.
3D rendering of a microcrack with its best-fitting ellipsoid.
The first two axis of the ellipsoid can be identified as the length and width of the microcrack.
Table 1.
3D quantitative data measured on the selected microcracks.
Table 2.
Microarchitecture and microdamage measurement on the five specimens.