micro-ct analysis of mouse long bones (proximal tibia, distal femur) · mouse long bones (proximal...

of 13

Micro-CT analysis of mouse long bones

(proximal tibia, distal femur)

Application note

of 13

2 Bruker-MicroCT method note: Long mouse bone micro-CT analysis

Introduction

The scans of mouse bones can be performed on the Skyscan 1072, 1172, 1076 micro-CT scanners. This report sets out a methodology for morphometric analysis of trabecular and cortical murine bone. Results are

presented for four mouse femur samples, for both trabecular and cortical bone morphometry.

Method

Scan and reconstruction parameters

The mouse bones were removed from alcohol storage and dried superficially on paper tissue, before being wrapped in plastic “cling-film”

or in parafilm, to prevent drying during scanning (and associated movement artefacts). Each plastic-wrapped bone was placed in a plastic / polystyrene foam tube which was mounted vertically in the Skyscan 1172

/ 1072, or horizontally in the 1076/1176 scanner sample chamber, for micro-CT imaging.

The suggested parameters of the scan in the 1172 (10/11 Mpix model) are as follows:

X-ray voltage 50 kV

X-ray current 200 µA.

Filter 0.5 mm aluminium

Image pixel size 4-5 µm

Camera resolution setting Medium or low (2000 or 1000 pixel field width)

Tomographic rotation 180/360

Rotation step 0.3-0.5

Frame averaging 1-2 (depending on required scan time)

Scan duration Medium res. 20-40 min.; low res. 6-10

min.

The suggested parameters of the scan in the 1172 (1.3 Mpix model) are

as follows:

X-ray voltage 50 kV




Camera resolution setting High (1280 pixel field width)


of 13



Frame averaging 1-2

Scan duration 30-50 minutes

The suggested parameters of the scan in the 1072 are as follows:

X-ray voltage 50 kV




Camera resolution setting n/a (1024 pixel field width)

Tomographic rotation 180

Rotation step 0.45

Frame averaging 1-2


The suggested parameters of the scan in the 1076/1176 (not in vivo, extracted bone) are as follows:

X-ray voltage 50 kV



Image pixel size 8.9 µm




Frame averaging 1


The suggested parameters of the scan in the 1076/1176 (in vivo, living

mouse under anaesthesia) are as follows:

X-ray voltage 50 kV



Image pixel size 8.9 µm


Tomographic rotation 180

of 13



Frame averaging 1


Reconstruction parameters suggested (same for all scanners):

Smoothing 0-1

Beam hardening 33%

Ring reduction 5

Post-alignment use value calculated by Nrecon

Histogram limits min 0; max – at right end of high density “tail”; refer to image to get good visible

contrast

Reconstruction was carried out employing a modified Feldkampi algorithm using the Skyscan Nreconii software which facilitates network distributed

reconstruction carried out on four pcs running simultaneously. Time for reconstruction of on scan dataset is usually much less than the scan

duration.

Region of interest selection for trabecular and cortical

bone

Both trabecular (metaphyseal) and cortical (metaphyseal-diaphyseal) were selected with reference to the growth plate. A crossectional slice was

selected as a growth plate reference slice, in the following way. Moving slice-by-slice toward the growth plate from the metaphysis/diaphysis, a point is reached where a clear “bridge” of low density cartilage

(chondrocyte seam) becomes established from one corner of the crossection to another. This bridge is established by the disappearance of

the last band of fine primary spongiosal bone interrupting the chondrocyte seam (see figure 1). This landmark allows a reference level to be defined

for the growth plate: trabecular and cortical volumes of interest are then defined relative to this reference level. (Note that the “bridge” can be at different locations in the growth plate crossection in difference bone

scans, depending on the orientation of the scanned bone.) Such a reference slice can be identified in all micro-CT scans of the proximal tibia

and distal femur.

Note that while in the femur there are four circular “islands” delineated by chondrocyte seams in crossections through the upper growth plate, in the

tibia there are only two such islands; none-the-less, the criterion for a growth plate reference described here can be reliably applied both for the

femur and tibia. The only possible exception is in very old mice or rats

of 13


where the growth plate is inactive and very little or no chondrocyte seams are present. In such cases, a similar landmark can be found involving

connection of the bony structures of the vestigial growth plate.

Figure 1. Moving slice-by-slice toward the growth plate from the metaphysis/diaphysis, a point is reached where a clear “bridge” of low

density cartilage (chondrocyte seam) becomes established from one corner of the crossection to another, as shown by the arrows. The two images above show the establishment of this bridge with the

disappearance of the last band of fine primary spongiosal bone interrupting the chondrocyte seam. This landmark allows a reference level

to be defined for the growth plate: trabecular and cortical volumes of interest are then defined relative to this reference level. (Note that the “bridge” can be at different locations in the growth plate crossection.)

The regions of interest (ROI) were selected with reference to a growth

plate reference slice, selected as described above. Trabecular and cortical regions were defined as positions along the long axis of the femur relative to the growth plate reference. The trabecular region commenced about

0.215 mm (50 image slices) from the growth plate level in the direction of the metaphysis, and extended from this position for a further 1.72 mm

(450 image slices). The vertical extent of the trabecular and cortical ROIs is indicated in figure 2.

Note that for the trabecular ROI the value of the number of slices of the

“offset” (interval from growth plate reference level to start of ROI, here 50 slices) and the “height” (here 450 slices) can be altered to allow for

different bone geometries in mice of different strain, age and gender. The offset needs to be sufficient so that the trabecular ROI begins in a region of mature secondary trabecular bone, without a significant presence of the

fine-structured growth-plate-associated primary spongiosa in the crossections.

Within the trabecular region, separation of the trabecular from cortical bone was done using a freehand drawing tool for delineating of complex

of 13


regions of interest (figure 3). The boundaries of the selected trabecular ROI ran parallel and close to the endocortical boundary – but excluding

peripheral vestiges of the growth plate and associated primary spongiosa, as shown in fig. 3. In Skyscan CT-analyser the volume of interest (VOI)

automatically interpolates or morphs the edited ROI shapes (such as freehand drawn shapes) between edited levels. The frequency of edited levels depends on how rapidly the bone crossection shape changes along

the bone axis (in the z direction) and becomes less as you move from the growth plate in the direction of the diaphysis (bone shaft).

The cortical region commenced about 2.15 mm (500 image slices) from the growth plate level in the direction of the metaphysis, and extended from this position for a further 0.43 mm (100 image slices). This probably

represents a diaphyseal site - there were relatively few remaining thin trabecular structures, compared to the much higher relative volume and

number density of trabecular structures at the metaphysis close to the growth plate. The trabecular and cortical bone ROIs were delineated as shown in figure 3.

Note again that for the cortical ROI the value of the number of slices of the “offset” (interval from growth plate reference level to start of ROI,

here 500 slices) and the “height” (here 100 slices) can be altered to allow for different bone geometries in mice of different strain, age and gender.

Figure 2. The vertical extent of the trabecular and cortical ROIs, defined with reference to a standard growth plate reference level (see text).

Growth plate (GP) reference level

Trab ROI: 0.215 – 1.94 mm from GP (50–450 slices); 1.72 mm section height (400 slices)

Cort ROI: 2.15 – 2.58 mm from GP (500–600 slices); 0.43 mm section height (100 slices)

of 13


Figure 3. Trabecular and cortical ROIs delineated by the Skyscan CT-

analyser software ROI tool, by freehand drawing.

Morphometric analysis

3D and 2D morphometric parameters were calculated for the trabecular

and cortical selected ROIs. Single grey threshold values were applied for the trabecular and cortical ROIs respectively (68 and 120). (The lower density threshold for trabecular bone allows segmentation of the finer

structures; the thicker cortical bone can have higher threshold to represent thickness and porosity more accurately.) The signal to noise

ratios of the reconstructed images were high so that further image processing steps on the binarised images such as despeckle were not likely to significantly improve measurement precision. 3D parameters

were based on analysis of a Marching Cubesiii type model with a rendered surface. Calculation all of 2D areas and perimeters was based on the

Prattiv algorithm. Morphometric parameters measured by CT-analyser have been validated on both virtual objects and aluminium foil and wire phantomsv. Structure thickness in 3D was calculated using the local

thickness or “sphere-fitting” methodvi, and structure model index (an indicator of the relative prevalence of plates and rods) was derived

according to the method of Hildebrand and Ruegseggervii. Degree of anisotropy was calculated by the mean intercept methodviii.

3D Model construction

Rendered 3D models were constructed for 3D viewing of trabecular and cortical analysed regions. Model construction was by the “Double time

cubes” methodix, a modification of the Marching cubes method3.

of 13


Results

Trabecular bone

Table 1 lists examples of measured morphometric parameters for

trabecular bone, within the volumes of interest referenced to the growth plate, as described above. Of the four examples A-D, samples A and B

have a lower trabecular percent volume (BV/TV), of 4-5%, while C and D have about double this percent volume, at 9-11%. Most of this increase in trabecular BV/TV is due to an increased spatial density of trabecular

structures, as indicated by trabecular number (Tb.N) being twice as high in C and D compared to A and B. By contrast the trabecular thickness

(Tb.Th) is not much different between the four samples.

Images of 3d rendered models of the four samples (the analysed trabecular volumes of interest) are shown in figure 4.

Cortical bone

The morphometric parameters measured for four cortical bone samples are listed in table 2. Samples C and D have greater cortical thickness and

crossectional area; however both periosteal and endosteal perimeter was greater in samples A and B. This implies that the cortical diaphysis was

both wider and thinner in A and B. The wider bone shape in A and B compensates for the reduced cortical thickness and area in terms of cortical strength, as indicated by rotational moment of inertia, which is

about the same in all four samples.

Images of 3d rendered models of the four samples (the analysed cortical

volumes of interest) are shown in figure 5.

of 13


Table 1. Trabecular morphometric parameters measured by micro-CT in mouse distal femur samples A-D.

Distal femur

Parameter Parameter symbolx, unit

Mouse A Mouse B Mouse C Mouse D

Total

(tissue) volume

TV (mm3) 3.1385 3.1942 2.3077 2.4298

Bone volume BV (mm3) 0.1249 0.1500 0.2510 0.2249

Percent bone volume

BV/TV (%) 3.9786 4.6965 10.8751 9.2577

Bone surface BS (mm2) 11.3041 12.4453 17.1408 16.8528

Bone surface to volume ratio

BS/BV (mm-1) 90.5295 82.9592 68.2993 74.9197

Bone surface density

BS/TV (mm-1) 3.6018 3.8962 7.4276 6.9358

Trabecular thickness

Tb.Th (mm) 0.0475 0.0505 0.0553 0.0514

Trabecular separation

Tb.Sp (mm) 0.3413 0.3507 0.2754 0.2842

Trabecular number

Tb.N (mm-1) 0.8380 0.9293 1.9678 1.8013

Trabecular pattern

factor

Tb.Pf (mm-1) 33.5844 30.8598 18.1766 18.4714

Structure

model index

SMI 2.4090 2.3749 1.7773 1.7549

Euler

connectivity

E.Con 138.0 160.5 204.0 194.0

Euler

connectivity density

E.Con.D

(mm-3) 43.9700 50.2473 88.3997 79.8420

Degree of anisotropy

DA 1.7354 1.2388 1.8672 1.7124

of 13


Table 2. Cortical morphometric parameters measured by micro-CT in mouse distal femur samples A-D.

Distal femur

Parameter Parameter symbol9, unit

Mouse A Mouse B Mouse C Mouse D

Cortical

thickness (3D)

Ct.Th (mm) 0.1921 0.1921 0.2154 0.2230

Cortical

crossectional thickness

(2D)

Ct.Cs.Th

(mm)

0.1683 0.1724 0.1964 0.2008

Cortical

periosteal perimeter

Ct.Pe.Pm

(mm)

5.1556 5.2004 4.9922 4.9303

Cortical endosteal perimeter

Ct.En.Pm (mm)

4.0622 3.9723 3.8091 3.6415

Cortical crossectional

area

Ct.Ar (mm2) 0.7754 0.7905 0.8643 8.5717

Polar

moment of inertia

MMI(p)

(mm4)

0.3586 0.3613 0.3540 0.3392

Eccentricity Ecc 0.7662 0.7542 0.7713 0.7591

Cortical

porosity

Ct.Po (%) 0.0846 0.0352 0.0709 0.0538

of 13


(a) (b)

(c) (d)

Figure 4. 3D rendered models from the trabecular ROIs from mouse femur samples A-D

of 13


(a) (b)

(c) (d)

Figure 5. 3D rendered models of the cortical ROIs from the distal femurs

of samples A-D.

of 13


i Feldkamp LA, Davis LC, Kress JW (1984) Practical cone-beam algorithm. J. Opt. Soc. Am. 1 (6):612-619.

ii Xuan Liu, Alexander Sasov (2005) Cluster reconstruction strategies for microCT and nanoCT scanners. In: Proceedings of the Fully Three-Dimensional Image Reconstruction Meeting in Radiology and Nuclear

Medicine, July 6-9, 2005, Fort Douglas/Olympic Village, Salt Lake City, Utah, USA.

iii Lorensen WE, Cline HE (1987) Marching cubes: a high resolution 3d surface construction

algorithm. Computer graphics 21 (4): 163-169.

iv Pratt WK. Digital image processing, 2nd ed. New York: Wiley; 1991. Chapter 16.

v Salmon PL, Buelens E, Sasov AY (2003) Performance of In Vivo Micro-CT Analysis of Mouse Lumbar Vertebral and Knee Trabecular Bone

Architecture. J. Bone Miner. Res.18 (suppl. 2) S256.

vi Hildebrand T and Ruegsegger P (1997) A new method for the model independent assessment of thickness in three dimensional images. J.

Microsc.185: 67-75.

vii Hildebrand T and Ruegsegger P (1997) Quantification of bone

microarchitecture with the structure model index. Comp. Meth. Biomech. Biomed. Eng. 1: 15-23.

viii Harrigan T P and Mann R W (1984) Characterisation of microstructural anisotropy in orthotropic materials using a second rank tensor. J. Mater. Sci. 19: 761-767.

ix Bouvier Dennis J (2004) Double-time cubes: a fast 3d surface construction algorithm for volume visualisation. University of Arkansas,

313 Engineering Hall, Fayetteville, AR72701, USA.

x Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, Ott SM, Recker RR (1987) Bone Histomorphometry: standardization of

nomenclature, symbols and units. J. Bone Miner. Res. 2 (6): 595-610.

micro-ct analysis of mouse long bones (proximal tibia, distal femur) · mouse long bones (proximal...

Documents