7.2 Daily Quality Control, Phantoms, and Control Charts
Key Takeaways
- Daily quality control (QC) scanning is mandatory on all central DXA systems prior to patient imaging to detect instrument drift, detector degradation, and x-ray source instability before diagnostic errors occur.
- Anthropomorphic spine phantoms simulate human lumbar vertebrae surrounded by tissue-equivalent resin with certified bone mineral density (BMD) values, while block step-wedge phantoms calibrate system linearity.
- The daily calibration sequence executes automated high-voltage ramping, detector dark current and background radiation checks, internal reference drum or wheel synchronization, and standardized phantom positioning.
- Shewhart (Levey-Jennings) control charts establish an operational baseline from at least 10 consecutive daily scans, defining Warning Limits at ±1.5% (±2 SD) and Action/Control Limits at ±3.0% (±3 SD).
- Out-of-control rules dictate repeating the scan immediately for warning limit violations, notifying service if consecutive scans exceed warning limits, and halting all patient scanning immediately if any scan breaches the ±3.0% action limit.
7.2 Daily Quality Control, Phantoms, and Control Charts
Quick Answer: Daily quality control (QC) is mandatory on all central DXA systems prior to scanning patients. Automated daily procedures detect instrument drift, x-ray tube instability, and detector degradation before diagnostic errors occur. Daily phantom scans are plotted on Shewhart (Levey-Jennings) control charts against a baseline mean established from at least 10 consecutive scans. Warning limits are set at ±1.5% (±2 SD) and Action limits at ±3.0% (±3 SD). Any action limit violation mandates an immediate halt to patient scanning until a service engineer recalibrates the system.
1. Rationale and Regulatory Imperatives of Daily QC
Dual-energy x-ray absorptiometry measures subtle bone mineral density (BMD) changes over time. Because annual biological bone changes average only 1% to 3%, central DXA systems require strict longitudinal stability. An undetected calibration drift of 1.5% can obscure genuine drug responses or falsely suggest bone loss, altering patient care.
Daily QC validates system calibration, detector linearity, and high-voltage stability before exposing patients to radiation. Organizations including the International Society for Clinical Densitometry (ISCD), the American College of Radiology (ACR), and state radiation control bureaus mandate daily QC execution and permanent documentation every clinical operating day.
2. Calibration Phantom Types and Physical Engineering
To ensure reproducibility, standardized calibration phantoms with stable physical properties are scanned:
- Anthropomorphic Spine Phantoms: Replicate human lumbar anatomy (L1–L4) using precision calcium hydroxyapatite [$\text{Ca}_{10}(\text{PO}_4)_6(\text{OH})2$] or aluminum vertebrae encased in tissue-equivalent epoxy resin ($Z{eff} \approx 7.4$). Each level features certified, factory-calibrated BMD ($\text{g/cm}^2$) and bone mineral content (BMC). These phantoms serve as the universal standard for longitudinal stability and daily QA.
- Block (Step-Wedge) Phantoms: Fabricated with stepped aluminum and polyethylene slabs of varying thicknesses. They verify detector linearity across clinical densities and test edge-detection algorithms.
- External Multi-Density Standards: Phantoms such as the European Spine Phantom (ESP) contain three discrete density inserts ($0.5$, $1.0$, and $1.5 \text{ g/cm}^2$) to standardize multi-center trials and cross-calibrate scanners across manufacturers.
2b. Types of Quality Control: Internal Versus External, Baseline Versus Longitudinal
The ARRT outline divides quality control into two paired distinctions, and both appear as exam items.
Internal Versus External Calibration
| Internal calibration | External calibration | |
|---|---|---|
| What it is | Reference standards built into the scanner itself — a rotating drum or wheel carrying air, bone, and tissue-equivalent samples, or an internal reference in the beam path | A separate physical phantom placed on the table and scanned by the technologist |
| When it runs | Automatically, continuously during acquisition on some systems, or as part of the power-up self-test | On the facility's QC schedule, typically daily before the first patient |
| What it corrects | Real-time drift in tube output, detector gain, and electronics during the scan itself | Nothing — it detects, it does not correct. It verifies that the whole system, including the internal calibration, is producing stable values |
| Who sees it | Largely invisible to the technologist | Plotted on the QC chart and acted on by the technologist |
The practical point is that internal calibration is a correction mechanism and external phantom QC is a verification mechanism. A scanner whose internal calibration is failing will still report values — the external phantom is how you find out they are wrong.
Baseline Versus Longitudinal Phantom Scanning
| Baseline phantom scanning | Longitudinal phantom scanning | |
|---|---|---|
| Purpose | Establish the reference mean and control limits for this scanner | Monitor day-to-day stability against that reference |
| When performed | At installation, after relocation, and after major service or component replacement | Every operating day, before patient scanning |
| How many scans | A series, commonly at least 10 consecutive scans and optimally 20 to 30 | One scan per day |
| What it produces | The center line ($\bar{x}$) and the baseline standard deviation from which warning and action limits are derived | A single plotted point |
The relationship between them is the whole basis of QC charting: longitudinal scans are meaningless without a valid baseline. This is why relocation, a major repair, or a scanner replacement forces a new baseline rather than continuing the old chart — the reference against which the daily point is judged no longer describes the machine.
3. The Daily Operational Calibration Sequence
Before patient appointments, the technologist executes the automated daily sequence:
- High-Voltage Ramp & Self-Test: Ramps tube potential (kVp) and current (mA) to verify filament emission, high-voltage stability, and electrical ripple.
- Detector Dark Current & Background Check: With the x-ray shutter closed, the detector array measures baseline electronic noise and ambient radiation, which are subtracted from subsequent scans.
- Internal Reference Calibration: Scanners perform pixel normalization. Rapid kVp switching systems (Hologic) use an internal rotating drum with air, bone, and tissue standards, updating calibration on every scan line. K-edge filtration systems (GE Lunar) use periodic internal reference checks.
- Standardized Phantom Acquisition: The anthropomorphic phantom is placed on the table, aligned along the longitudinal centerline using laser lights, leveled to eliminate tilt, and scanned using standard parameters.
4. Shewhart and Levey-Jennings Quality Control Charts
Phantom BMD measurements are plotted chronologically on a Shewhart (Levey-Jennings) control chart:
- Baseline Center Line ($\bar{x}$): Established after installation or major service using the mean BMD of at least 10 consecutive daily phantom scans (optimally 20 to 30) under stable conditions.
- Baseline Standard Deviation ($\sigma$): Calculated directly from the baseline scan series to quantify baseline instrument noise.
- Warning Limits (UWL / LWL): Established at $\pm 1.5%$ of the baseline mean (or $\pm 2$ SD).
- Action / Control Limits (UAL / LAL): Established at $\pm 3.0%$ of the baseline mean (or $\pm 3$ SD).
5. Statistical Pattern Interpretation: Random Variation, Trends, and Shifts
Control chart evaluation distinguishes random scatter from systematic hardware failure:
Normal Random Variation
Daily points fluctuate symmetrically above and below the center line within $\pm 1.5%$ warning limits due to acceptable quantum noise and temperature fluctuations.
Systematic Trends (Drift)
Defined as 5 to 7 consecutive daily points moving progressively in one direction. Trends reflect gradual degradation: tube target aging, tungsten pitting, filter oxidation, high-voltage drift, or optical dust.
- Systematic Shifts (Step-Function Changes): Defined as 3 or more consecutive points falling entirely on one side of the center line, or an abrupt jump. Shifts indicate sudden mechanical or electronic changes: detector channel failure, optical misalignment, collimator slippage, or uncalibrated service.
6. Interpreting QC Results: Pass/Fail Criteria, Need for Repeat QC, and Need for Service
The ARRT outline breaks interpretation of QC results into three decisions, and the action table below answers all three for any given daily point:
- Pass/fail criteria — does the plotted point fall inside the warning limits, between the warning and action limits, or beyond the action limits?
- Need for repeat QC tests — a point between the warning and action limits is repeated immediately after re-seating and re-levelling the phantom, because a single out-of-range point is more often a phantom placement error than a scanner fault.
- Need for service — a point beyond the action limits, a failed repeat, a sustained trend, or a persistent shift all mean the scanner requires manufacturer service before patient scanning resumes.
Facilities must adhere to strict operational protocols based on daily phantom results:
| QC Deviation Category | Control Chart Parameter | Immediate Technical Action | Clinical Scanning Status |
|---|---|---|---|
| Normal Status | Within $\pm 1.5%$ of mean ($\le \pm 2$ SD) | Log result in QC archive | Proceed with clinical scanning |
| Warning Limit Violation | Between $\pm 1.5%$ and $\pm 3.0%$ (2 to 3 SD) | Reposition phantom, verify level, repeat scan immediately | If repeat is within $\pm 1.5%$, proceed; if repeat fails or occurs on 2 consecutive days, notify service |
| Action Limit Violation | Exceeds $\pm 3.0%$ of mean ($> \pm 3$ SD) | Abort workflow; do NOT scan patients; open service ticket | HALT all clinical scanning immediately until engineer completes corrective service |
| Systematic Trend | 5 to 7 points moving unidirectionally | Review tube logs; check filters; notify physicist | Scanning permitted under close monitoring; schedule preventive maintenance |
| Systematic Shift | Abrupt jump persisting $> 3$ scans | Verify software/hardware status; run diagnostic self-test | Scan duplicate phantom; contact service before patient scanning if shift persists |
7. Maintenance Records and Regulatory Compliance
QC documentation must be archived in a permanent facility logbook or digital audit archive:
- Ambient Room Parameters: Record daily room temperature (65°F–75°F / 18°C–24°C) and relative humidity (20%–80% non-condensing) to avoid detector drift.
- QC Logs: Store dated Levey-Jennings plots with technologist initials.
- Service History: Archive all service reports, component replacements, and recalibration datasets for 3 to 5 years (or equipment lifetime per state regulations).
On a dual-energy x-ray absorptiometry Levey-Jennings quality control chart, what statistical deviation threshold represents the action (control) limit that mandates an immediate halt to all clinical patient scanning?
What primary operational characteristic distinguishes a systematic trend from a systematic shift on a densitometry daily quality control chart?
A densitometry technologist performs the daily anthropomorphic spine phantom scan, and the software reports a bone mineral density measurement that deviates from the baseline mean by +1.8%. What is the mandatory immediate protocol?