4.3 High-Resolution Peripheral QCT & Emerging Technologies
Key Takeaways
Nominal voxel size is the reconstructed grid spacing. Measured spatial resolution also depends on the scan and reconstruction.
HR-pQCT reports compartment density and estimates of segmented structure. First- and second-generation analyses can differ.
Peripheral micro-FEA estimates stiffness or failure load under specified assumptions. Validated central QCT strength thresholds are method- and site-specific.
REMS reports ultrasound-derived estimates. Its clinical role depends on the device labeling and applicable guideline.
Read the scan protocol before comparing bone structure
High-resolution peripheral quantitative computed tomography (HR-pQCT) images the distal radius or tibia. The operator immobilizes the limb, uses a scout image to identify a reference landmark, and places the scan stack according to the study or device protocol. If a repeat scan samples a different anatomical segment, an apparent change can arise from the offset alone. Check motion, scan coverage, positioning and segmentation before interpreting an apparent difference.
Standard XtremeCT and XtremeCT II patient protocols have nominal isotropic voxel sizes of about 82 µm and 61 µm respectively. Voxel size is the spacing of the reconstructed grid. The smallest feature that can be distinguished depends on the acquisition and reconstruction as well as contrast and the method used to characterize resolution. In one experimental comparison, resolution at 10% modulation transfer was about 142 µm on the original scanner and 95 µm on the second-generation scanner with their standard settings. Partial volume and motion can obscure small features even when the grid spacing is smaller. The manufacturer reports about 3 µSv and 5 µSv for example protocols on the respective systems. Different scan lengths and settings can change the dose.
Distinguish density from segmented structure
| Output | How to interpret it |
|---|---|
| Trabecular or cortical volumetric BMD | A calibrated density estimate for the corresponding compartment, usually in mg hydroxyapatite equivalent/cm³. Cortical compartment density depends on which voxels and pores are included. |
| Trabecular bone volume fraction | Bone volume divided by the analyzed trabecular tissue volume. In the first-generation standard method, it is estimated from trabecular vBMD using 1,200 mg hydroxyapatite equivalent/cm³ as the assumed fully mineralized density. The standard second-generation approach uses segmented bone volume. |
| Trabecular thickness and separation | First-generation standard values use derived estimates. The higher-resolution second-generation protocol supports image-based three-dimensional morphometry. Values remain dependent on the analysis. |
| Cortical thickness and porosity | These describe the segmented cortex. Estimated porosity divides the volume labeled as pores by the combined solid and pore volume. Small pores and partial-volume voxels may be missed. |
An analyst should record the scanner generation, site and segmentation method. A higher thickness result after changing scanners could reflect the analysis method. An in vivo cross-calibration study found that thickness was harder to reproduce across generations than many density measurements. Check validated comparability before interpreting the change as biological.
Interpret diabetes findings at the study level
Type 2 diabetes can raise fracture risk even with preserved areal BMD. In one cohort of women with diabetes, tibial cortical porosity was higher on average than in controls, while estimated failure load did not differ between those two groups. Within the diabetes group, women with a history of atraumatic fracture had several cortical and trabecular deficits and lower estimated failure load than women without fracture. Other findings vary by site and subgroup. A peripheral scan does not determine the cause of an individual's fracture.
Interpret a finite element result
Micro-finite element analysis converts the segmented peripheral image into a mesh, assigns material properties and applies specified virtual loads. Apparent stiffness is reaction force divided by imposed displacement. A model yielding 2.4 kN at 0.30 mm displacement has an apparent stiffness of 8 kN/mm under those conditions. The failure-load estimate depends on the model and its failure criterion. It is a simulation output, so changing model assumptions can change the result even with the same image.
BoMIC followed 7,254 participants and found associations between peripheral failure-load estimates and incident fracture. In models adjusted for demographic, cohort and body-size factors, hazard ratios per one standard deviation lower estimated failure load were 2.40 at the tibia and 2.13 at the radius. Associations attenuated after additional adjustment for femoral neck BMD or FRAX. These findings do not create an individual fracture probability or a universal treatment cutoff for HR-pQCT.
For a separate clinical use, ISCD permits validated central QCT finite element estimates of femoral or vertebral strength to inform treatment with clinical risk factors. For example, the FOCUS study used method-specific fragile femoral strength cutoffs of 3,000 N for women and 3,500 N for men. These are strength cutoffs for that implementation. They do not transfer to peripheral micro-FEA, and neither strength estimate supplies a WHO T-score.
Recognize another imaging approach
Radiofrequency echographic multispectrometry (REMS) uses ultrasound echoes from the lumbar spine or femoral neck to estimate device-specific density-related measures. An FDA-cleared EchoS Family implementation can aid assessment alongside other clinical risk factors. Its Fragility Score module estimates fracture risk. Ultrasound uses no ionizing radiation, but image quality, acquisition and applicable instructions still matter. The UK NOGG guideline does not recommend REMS for diagnosing osteoporosis. A software density estimate does not measure the structure of individual trabeculae or guarantee immunity to degenerative findings.
SCANCO patient protocols · SCANCO protocol dose example · Resolution validation · HR-pQCT assessment guidance · BoMIC · Diabetes cohort · FOCUS · ISCD Adult Official Positions · FDA EchoS Family summary · UK NOGG guidance
A 60.7 µm reconstructed voxel is reported for an HR-pQCT scan. What does this number establish?
Every cortical pore wider than 60.7 µm is resolved reliably.
It is the nominal grid spacing. Feature detection also depends on spatial resolution, contrast and analysis.
It is the minimum detectable BMD change between different scanners.
It is the maximum trabecular thickness at this site.
A peripheral HR-pQCT micro-FEA simulation gives a reaction force of 2.4 kN at a displacement of 0.30 mm. What can be concluded?
Apparent stiffness is 0.125 kN/mm and establishes osteoporosis.
Failure load must be 8 kN in every loading direction.
Apparent stiffness is 8 kN/mm under the stated model conditions.
The result can be compared with a validated hip CT strength cutoff to classify the radius.
An HR-pQCT cohort found higher mean tibial cortical porosity in women with type 2 diabetes than in controls, but no difference in estimated failure load between those groups. Which interpretation is supported?
The findings describe a site- and cohort-specific association. They do not establish one mechanism for all diabetic fractures.
Every patient with type 2 diabetes has increased radial porosity.
Every patient with an osteoporosis-range DXA score has low tibial failure load.
The group comparison proves that higher porosity can never affect mechanical behavior.
Which statement correctly describes radiofrequency echographic multispectrometry (REMS)?
It counts individual trabeculae directly in a central X-ray CT image.
It is another name for heel broadband ultrasound attenuation.
Its density result has the same WHO diagnostic status in every jurisdiction.
It analyzes ultrasound echoes at central sites to produce device-specific estimates. Apply the relevant clinical guidance.
Sections you finish are checked off in the contents.