5.4 Technical Artifacts & Pitfalls in DXA Interpretation
Key Takeaways
Because DXA measures a projection, superimposed structures and analysis errors can change apparent BMD.
Degenerative sclerosis, vertebral compression, hardware, contrast and tablets can affect spine measurements. Inspect the image to determine whether the affected ROI is valid.
Laminectomy can lower projected mineral at the operated level. A localized change or artifact does not establish generalized bone loss.
Correct a recoverable analysis problem and keep all evaluable sites and vertebrae. Document results that cannot be interpreted.
Inspect the image before interpreting the value
Dual-energy X-ray absorptiometry (DXA) reports areal bone mineral density as bone mineral content (BMC) divided by projected bone area. An image includes all structures in the beam path. The software estimates how much of the attenuation comes from bone and uses surrounding soft tissue in its analysis. An overlapping object, a change in anatomy, or an error in identifying the bone can affect the estimate. The magnitude and direction depend on where the finding lies and how the software handles it.
Inspect the images with the clinical history and check the bone map and regions of interest (ROIs) before accepting unusually high or low values. In particular, compare the vertebrae with neighboring levels and the prior image. A marked discrepancy flags a technical or structural question. It does not tell you the cause by itself.
| Finding | Potential effect | Interpretation response |
|---|---|---|
| Focal osteophytes, facet sclerosis or ligament ossification | Additional mineral in the projection can elevate spine BMD | Review the relevant levels and retain all that are evaluable |
| Vertebral compression | Reduced projected area and trabecular impaction may elevate areal BMD | Assess the altered anatomy and exclude the level if non-assessable under the ISCD BMD rules |
| Hardware, vertebral augmentation cement or mineralized soft tissue over the ROI | May contaminate the bone map or soft-tissue estimate | Identify the location and decide whether a valid analysis is possible |
| Barium, other radiopaque contrast or an undissolved calcium tablet | Can interfere with the bone or surrounding soft-tissue measurement | Review the scan and preparation history, then follow local protocol for repeat imaging if needed |
| Laminectomy or a lytic lesion | Removal or loss of local bone can lower projected mineral measurement | Consider local alteration and any need for clinical or imaging evaluation |
| External metal, motion, soft-tissue folds or an incorrect ROI | Can alter edge detection, projected area, or soft-tissue correction | Correct the cause or repeat as appropriate when diagnostic quality is inadequate |
Dense material is often a source of falsely high values. In some circumstances it can produce the opposite effect, such as when very dense metal is excluded by the software and appears as a gap in the bone map. Review the raw appearance and analysis together. The value from a lytic lesion or a laminectomy level can also be low because that region has changed structurally. It should not be treated as a representative measurement of generalized bone density.
Separate correctable analysis from invalid anatomy
- Confirm patient details, indication and history of procedures, contrast or supplements that could affect the result.
- Inspect coverage, patient positioning, motion and the projection. Compare with the prior images for a serial study.
- Confirm landmarks and ROI boundaries. An intervertebral marker drawn through bone can misassign projected area and BMC between adjacent levels. Review the bone map for missed bone and inclusion of non-target structures.
- Correct the analysis using validated tools if the image permits it. For a correctable acquisition problem, repeat the affected measurement when appropriate. If the anatomy or artifact makes a site unusable, document the reason and keep valid results at other sites.
- In the spine, retain all evaluable vertebrae. If an artifact affects only L2, the eligible L1, L3 and L4 levels can be used. If only one level remains, use another valid site for diagnosis. Match the remaining levels and threshold before assessing serial change.
Worked examples
A patient who just took calcium carbonate has an unexpected dense object overlapping L2. The other levels look normal on the scan. The object can contaminate the L2 measurement, so assess whether the scan can be repeated appropriately or L2 must be excluded. A result from L1, L3 and L4 can still contribute to diagnosis if those levels are technically valid. Follow the facility's preparation instructions and review the images again. There is no need to assume a specific delay or order a radiograph solely to satisfy a universal DXA rule.
A different patient has a wedge compression fracture and an unexpectedly high BMD at the affected level. If projected area decreases proportionally more than the measured BMC, their ratio increases. This measurement can reflect compaction and geometry. It does not establish that the fractured vertebra is stronger. Evaluate technical validity and use the intact levels where possible.
A third patient's L3 value is lower than its neighbors following decompressive laminectomy. Missing posterior elements can lower the measured mineral in the PA projection. Review the surgical history and images to distinguish this local cause from a generalized skeletal finding. If L3 is non-assessable, exclude it and document the reason.
ISCD Adult Official Positions · International Working Group DXA practice guideline · ISCD tablet artifact case · ISCD laminectomy case
A follow-up PA lumbar DXA shows a dense object directly over L2. The patient took calcium carbonate shortly before the scan. L2 is contaminated, but L1, L3 and L4 are evaluable. What is the appropriate action if the artifact cannot be removed from the analysis?
Report all four levels because an oral tablet cannot affect the projection.
Omit L2 from the analysis and use the valid L1, L3 and L4 levels. Consider a repeat acquisition if appropriate.
Multiply L2 BMD by an estimated calcium correction factor.
Keep only L3 and L4 because valid levels must be contiguous.
A PA lumbar DXA shows a wedge compression fracture with an unexpectedly high areal BMD at the affected level. Which change in the projection could explain the high value?
Hematoma always increases the soft-tissue ratio by a fixed amount.
Callus within 48 hours quadruples volumetric BMD.
Projected vertebral area falls proportionally more than the measured BMC, increasing their ratio.
The fracture converts the DXA measurement to a volumetric density.
A patient previously underwent extensive decompressive laminectomy at L3. Which finding can occur at the operated level on PA lumbar DXA?
An obligatory rise in BMD from scar tissue calcification.
A higher value because the surgery always shifts both transverse processes.
Automatic merging of the level into the L4 bone map.
Reduced projected BMC and a lower areal BMD because posterior elements are absent.
Sections you finish are checked off in the contents.