7.5 Maintenance, Repair, Relocation, Cross-Calibration, and Software Upgrades

7.5 Maintenance, Repair, Relocation, Cross-Calibration, and Software Upgrades

Quick Answer: Anything that changes the scanner can change the number. Relocation requires recalibration and a fresh phantom baseline. Service or component replacement requires QC verification before patients resume. Replacing the system requires cross-calibration — at least 10 phantom scans on each unit for the same make and model, investigating a difference greater than 1%; approximately 30 patients scanned on both for a different make or model. Software upgrades can change reported T-scores without any change in the patient.

The Underlying Rule

Serial densitometry rests on one assumption: that the measurement system today is the same measurement system as at baseline. Every topic in this section is a way that assumption breaks, and every response is a way of restoring it or of honestly declaring it broken.

Relocation

Moving a densitometer — to a new room, a new building, or onto a mobile coach — disturbs mechanical alignment, and it changes the surrounding structural and environmental conditions.

Required sequence after any relocation:

1. Manufacturer service installation and mechanical alignment. The scan arm, source, and detector array must be re-aligned and verified by qualified service personnel.
2. Recalibration. The system's internal calibration is re-established at the new location.
3. Re-establish baseline phantom data. The prior phantom baseline mean and control limits were derived at the old location on the old alignment and are no longer valid. A new baseline is created from a fresh series of phantom scans — typically at least 10 — and new control limits are computed from it.
4. Regulatory and physicist requirements. State radiation control regulations commonly require notification of a change of location and a radiation survey of the new room before use.
5. Document the date and nature of the move in the QC record, so future serial comparisons crossing that date can be evaluated.

Patient scanning does not resume until the new baseline is established and in-tolerance.

Service, Repair, and Component Replacement

Routine preventive maintenance follows the manufacturer's schedule. Unscheduled service follows a QC failure, a mechanical fault, or an image artifact.

The technologist's obligations around service:

• Recognize and escalate. Daily QC that fails, trends toward a control limit, or shows a sudden step is a service trigger. Repeated QC failures are not "noise to scan through."
• Do not scan patients on an out-of-tolerance system. A scanner failing QC produces values that cannot be trusted for diagnosis or monitoring.
• Verify QC after service. Any service touching the x-ray source, detector, collimator, scan arm mechanics, or calibration requires QC verification before patients resume, and often a new phantom baseline when a major component is replaced.
• Keep the service log. Dates, the fault, what was replaced, and the post-service QC result. When a patient's follow-up scan shows an unexpected shift, the service log is the first place to look.

Cross-Calibration

Cross-calibration answers the question: "Can values from the new system be compared numerically to values from the old one?"

Two consequences follow directly. First, a new precision study is required for each technologist on the new system, because precision is system-specific and the existing least significant change does not transfer. Second, cross-calibration is a facility-level responsibility involving the physicist and manufacturer, not something a technologist performs alone — but the technologist is usually the person who notices that a patient's baseline was acquired elsewhere and asks the question.

Note that cross-calibration appears as a named topic in the content specifications in effect through December 31, 2026. It is removed as a named item in the specifications effective January 1, 2027, but the underlying practice remains clinically essential, and the principle that values are not comparable across different systems is examined in both versions through the follow-up-scanning and serial-comparison topics.

Software Errors, Upgrades, and Updates

Software changes are the least visible and most under-appreciated source of apparent change.

• Analysis algorithm updates can alter edge detection and ROI defaults, changing reported BMC, area, and BMD on the same raw data.
• Reference database updates change T-scores and Z-scores without changing measured BMD at all. A patient can appear to have improved or worsened when nothing physical happened.
• Archived scans should be re-analyzed on the new version where the manufacturer advises it, so that baseline and follow-up are analyzed by the same algorithm.
• Record the software version with each study. A version change is a discontinuity in the record.
• Software errors — failed analyses, corrupted files, crashes during acquisition, failed archiving — are reported to the manufacturer and logged. Do not repeat a patient scan to work around a software fault until it is clear the acquisition, and not the analysis, failed.

This is the strongest practical argument for the rule stated throughout this guide: compare absolute BMD in g/cm², not T-scores. BMD is a physical measurement; T-scores are the output of a database that the vendor may change.

Reconfiguring Report Settings

The specifications effective January 2027 add reconfiguration of report settings — for example, enabling or disabling automatic analysis — as a named topic.

• Automatic analysis speeds throughput but can accept a poor ROI or a bad edge map without review. Facilities that disable auto-analysis force explicit technologist review of every ROI.
• Default site and ROI settings, displayed regions, reference database selection, and which values print on the report are all configurable, and changing them changes what clinicians see.
• Configuration changes must be deliberate, documented, and facility-wide, not made ad hoc by an individual technologist mid-series. An ROI default silently changed between a baseline and a follow-up creates an analytic difference that looks biological.

Equipment Safety

• Emergency stop. Every scanner has an emergency stop control that halts scan arm motion and terminates the exposure. Know where it is on your system, verify it is unobstructed before each session, and never disable or obstruct it. Use it for any risk of patient entrapment or injury by the moving arm.
• Electrical safety. Inspect power cords and connections for damage; keep liquids away from the console and gantry; do not use extension cords or adapters on medical equipment; ensure the unit is on its designated circuit with intact grounding. Report any tingling, burning odor, sparking, or breaker trip immediately and remove the unit from service.
• Mechanical safety. Confirm the patient's limbs, gown, blankets, IV lines, and oxygen tubing are clear of the scan arm path before starting, and observe the patient throughout arm travel.
• Table weight limits. Know the stated limit. Exceeding it risks mechanical failure with a patient on the table.