3.3 Instrument Quality Control, Phantoms & Calibration
Key Takeaways
Follow the scanner manufacturer’s maintenance guidance. The ISCD Adult Positions advise at least weekly phantom scans when the manufacturer protocol does not already specify them.
Plot phantom BMD and review the series for outliers, sustained shifts or drift. The QC program sets corrective-action thresholds and records the response.
Check phantom mean BMD after service. The equipment change determines which cross-calibration procedure applies to serial patient measurements.
Verify the instrument before interpreting patient change
Quality control (QC) tracks whether the DXA system produces consistent measurements. A phantom supplies a stable object with fixed geometry and attenuation characteristics. Scanning it repeatedly over time can reveal a shift, gradual drift, or increased variability. Use the phantom, positioning procedure, scan mode, and limits specified for the instrument. Even stable phantom readings cannot establish clinical precision. The clinical precision study measures the variability introduced when real patients are scanned and repositioned.
The International Society for Clinical Densitometry (ISCD) directs facilities to follow the manufacturer's maintenance guidance. Its additional QC recommendations, when the manufacturer protocol does not already provide them, include periodic phantom scanning at least weekly, reviewing plots of phantom and calibration results, and verifying mean phantom BMD after service. Maintain service logs, set corrective-action thresholds that trigger a service call, and meet applicable inspection and radiation-survey requirements. If the manufacturer's instructions call for daily QC before scanning patients, follow that schedule.
Read the QC history
A control chart places phantom BMD measurements in date order against a reference mean and the limits specified in the QC program. Review the latest result against those limits and inspect the pattern of recent results. Use comparable acquisitions when establishing or reviewing the reference. Mixing different phantoms or protocols can shift the plotted values independently of scanner performance.
| Pattern | What to check | What to do |
|---|---|---|
| One outlying scan | Phantom identity, position, scan mode, recent QC history, and instrument messages | Follow the local corrective-action procedure, including repeat QC when indicated |
| Measurements moving gradually in one direction | Maintenance history, protocol consistency, and evidence of instrument drift | Compare the series with the program's action criteria and escalate as required |
| An abrupt shift after service | Phantom mean before and after the intervention | Verify the post-service mean and follow the applicable cross-calibration procedure |
| Repeat QC fails the action criterion | The repeat result and the manufacturer's troubleshooting instructions | Hold patient acquisition when required by the QC protocol and seek service |
A run of measurements on the same side of the reference mean can raise concern even if each point is within the single-scan limit. It does not identify the failed component. Determine whether there was a change in phantom positioning, scan protocol, or equipment and review the pattern under the facility's rule. ISCD calls for setting and enforcing action limits. The published Adult Positions leave the operational limits and run rules to the applicable program.
The average of the phantom readings is their sum divided by the number of readings. A chart may show the spread of the readings with a sample standard deviation, which uses the sum of squared deviations from the sample mean divided by one fewer than the number of readings, followed by a square root. Whether a departure warrants intervention depends on the relevant QC rule. A reference mean and a statistical control limit are different from a patient's least significant change.
Worked QC response
Suppose the facility's approved procedure flags the first phantom reading of the day. Start by checking that the phantom and scan protocol match the intended QC test. Reposition and repeat if the procedure calls for it. If the repeat still fails the action threshold, follow the hold and service instructions in that procedure. Document both results and the eventual corrective action. A technologist who changes the center line simply to bring a failed result inside the chart conceals the evidence needed to evaluate scanner stability.
Now consider eleven consecutive scans below the reference mean, all within the single-measurement limits. Review this sustained pattern under the local trend criteria and check the service and acquisition records. If the trend meets the facility's action criterion, follow the escalation procedure even though no individual point is an outlier. An isolated result and a sustained run are different signals.
Document service and verify comparability
Record the QC result, failed or repeated checks, service date, equipment changes, and the verification used before returning the unit to clinical use. This documentation helps the interpreter identify which patients' serial measurements straddle a change in equipment. Verify the phantom mean after service. A hardware change or a system replacement can require cross-calibration to support longitudinal comparison. A documented change in phantom BMD alone does not justify subtracting that difference from patient measurements.
The facility’s QC procedure calls for an immediate repeat after a failed phantom check and a hold on patient scanning if the repeat fails. The first morning scan exceeds the action limit. What should the technologist do?
Replace the X-ray tube immediately based on the first reading.
Shift the control-chart mean to the new measurement and start patient scanning.
Check phantom setup and repeat the scan. If the repeat fails, follow the hold and service procedure.
Proceed with patient scanning because the first failed check cannot reflect scanner instability.
Eleven consecutive phantom readings fall below the reference mean, although each reading is within the single-scan limits. How should the densitometrist interpret the pattern?
Ignore it because no single scan exceeded the action limit.
Diagnose catastrophic detector failure from the run alone.
Switch to a different phantom before reviewing the trend.
Evaluate the run under the facility’s trend criteria and review acquisition and service records.
A service engineer replaces the X-ray tube on an existing DXA scanner. Which procedure addresses verification of phantom BMD across the hardware change?
Compare the mean from ten repositioned phantom scans before service with ten after service and contact the manufacturer if the difference is greater than 1%.
Resume routine patient scans immediately because the scanner model is unchanged.
Set the reference mean from a single post-service measurement and erase the earlier QC records.
Apply the difference from the first post-service reading directly to all patient BMD values.
Sections you finish are checked off in the contents.