Quality Assurance and Quality Control in Medical Imaging

Overview and Principles of Quality Assurance

Quality assurance (QA) in medical imaging encompasses the entire system of management policies, procedures, and processes implemented to ensure that every imaging examination is appropriate, technically optimal, correctly interpreted, communicated in a timely manner, and delivered at the lowest practicable radiation dose. QA is broader than quality control (QC): it includes clinical governance, credentialing, continuing medical education, protocol optimisation, equipment maintenance, preventive maintenance, calibration, and performance monitoring across the full imaging chain.

Quality control is the subset of QA concerned specifically with the technical performance of equipment. QC involves systematic, periodic testing of imaging systems against defined standards to identify drift, degradation, or failure before these translate into suboptimal clinical images or patient harm.

Core Principles of QA

Benefits of QA

• Patient safety: Detection of equipment malfunction before patients are harmed by poor images requiring repeat exposures or misdiagnosis
• Dose optimisation: Benchmark doses against Diagnostic Reference Levels (DRLs); DRLs are set at the 75th percentile of surveyed institutional doses and prompt investigation when exceeded (achievable doses are typically set lower, as an optimisation target)
• Image quality: Ensures spatial resolution, contrast resolution, noise, and uniformity remain within specified tolerances
• Regulatory compliance: Compliance with national and international standards (ARPANSA, IAEA, ACR accreditation programmes)
• Medicolegal protection: Documented evidence of system performance
• Continuous improvement: Trend analysis enables proactive maintenance rather than reactive repair

Increased QA Requirements for Asymptomatic (Screening) Populations

Screening programmes differ fundamentally from diagnostic imaging in that they are applied to large populations of healthy, asymptomatic individuals where the pre-test probability of disease is low. This has critical implications for the required QA framework.

Why Stricter QA Is Needed in Screening

1. Narrower harm-to-benefit ratio: Asymptomatic individuals receive no immediate therapeutic benefit; any avoidable dose, false positive, or missed lesion represents a net harm.
2. High volume, low prevalence: Even small degradations in sensitivity or specificity have large population-level consequences - increased false negatives yield missed cancers; false positives yield unnecessary biopsies and anxiety.
3. Minimum detectable lesion size is critical: Screening depends on detecting early-stage, often small lesions (e.g., microcalcification clusters, architectural distortion in mammography); equipment not performing to specification will miss these.
4. Regulatory requirements are heightened: Mammography screening is one of the most stringently regulated imaging applications. In the United States, the Mammography Quality Standards Act (MQSA, enacted 1992, CFR Title 21 Part 900) makes many ACR QA procedures mandatory, making mammography the only federally regulated imaging modality in the US. Accreditation and certification by the FDA are required before a facility may provide mammography services.
5. Programme-level auditing: Recall rates, cancer detection rates, positive predictive values, and interval cancer rates must be tracked as QA outcome measures beyond equipment testing.
6. Clinical image quality feedback is mandated: In mammography, the FDA's EQUIP initiative (launched January 2017) requires the Lead Interpreting Radiologist to provide ongoing feedback on image quality, document corrective actions, and conduct regular technologist image quality reviews.

Screening-Specific QA Measures

For mammography, QC tests are categorised under the regulatory framework:

• Category A tests (equipment must not be used for patient imaging until resolved): system resolution, breast dose, phantom image quality, SNR/CNR, flat-field calibration, compression paddle operation
• Category B tests (failed device cannot be used for interpretation, but imaging may continue on alternate approved devices): review workstation and laser film printer performance
• Category C tests: less time-critical performance checks

For digital mammography and digital breast tomosynthesis (DBT), manufacturers' QC procedures must be followed; an FDA-approved manufacturer-independent harmonised QC programme is available through the 2018 ACR Digital Mammography Quality Control Manual.

Quality Control Testing by Imaging Modality

Radiographic and Fluoroscopic Equipment QC

Radiographic QC is performed at acceptance testing, after any major repair, and at scheduled intervals (typically annually by a medical physicist, with more frequent technologist-level checks).

The half-value layer is defined as:

$$\text{HVL} = \frac{\ln 2}{\mu}$$

where $\mu$ is the linear attenuation coefficient of the attenuating material (typically aluminium). HVL measurement verifies beam hardening and filtration adequacy; inadequate filtration increases patient dose without diagnostic benefit.

Spatial resolution testing uses line-pair phantoms or star patterns. The observer identifies the limiting spatial frequency (reported in $\text{lp} \cdot \text{mm}^{-1}$) as the smallest resolvable bar group. For fluoroscopic systems, additional tests include dose rate, frame rate, and automatic brightness control performance.

Reject/retake analysis is an important ongoing QC component: portable chests, lumbar and thoracic spines, KUBs, and abdomens typically have the highest retake rates. Digital receptors (photostimulable phosphor plates or direct digital detectors) substantially reduce retake rates compared with film-screen systems.

Nuclear Medicine Equipment QC

Gamma Camera (Planar and SPECT)

QC of the Anger scintillation camera is essential because non-uniformity, non-linearity, and energy resolution errors introduce systematic artefacts into planar and tomographic images. In SPECT, even small (1-2%) non-uniformities are amplified into ring artefacts by the reconstruction process; flood correction maps must therefore be updated regularly.

Dose Calibrator QC

The dose calibrator assay is often the only check that a patient receives the prescribed radiopharmaceutical activity; its QC is therefore mandated by the NRC and state agencies, in accordance with nationally recognised standards or the manufacturer's instructions.

Note: attenuation correction factors for specific radionuclides, containers, and dose calibrator models may be required for accurate assay; this is an unresolved challenge for some radionuclides.

PET and PET/CT QC

Acceptance testing follows the NEMA NU2-2018 standard, which defines methodology for: 1. Spatial resolution 2. Scatter fraction, count losses, and randoms 3. Sensitivity 4. Accuracy of corrections for count losses and randoms 5. Image quality, accuracy, and corrections 6. Time-of-flight (ToF) resolution (where applicable) 7. PET/CT co-registration accuracy (hybrid systems)

Results are compared against manufacturer specifications.

Routine testing schedules are additionally specified by AAPM Report No. 126 (PET/CT Acceptance Testing and QA, 2019) and individual manufacturer recommendations.

ACR PET accreditation phantom - three sections:

All measurements are made on images reconstructed with full corrections applied (attenuation, scatter, randoms, dead time).

MRI Equipment QC

MRI QC must cover the entire imaging chain: the main magnetic field ($B_0$), radiofrequency (RF) system, gradient system, receiver coils, and display systems. The ACR MRI accreditation programme specifies minimum QC requirements, assessing personnel qualifications, equipment performance, QC procedure effectiveness, and clinical image quality.

Key components requiring periodic monitoring: - Magnetic field strength and homogeneity (shimming effectiveness) - Gradient linearity and calibration (affects geometric accuracy and slice profile) - RF tuning and transmitter gain (affects flip angle accuracy and SNR) - Receiver coil impedance matching and performance - Environmental noise sources and power supplies

ACR MRI accreditation phantom: cylindrical vessel, 190 mm internal diameter × 148 mm internal length, filled with a nickel sulphate/sodium chloride aqueous solution adjusted to provide clinically relevant T1 and T2 relaxation times. Contains geometric arrays for spatial accuracy, resolution inserts, and structures for slice position and thickness verification.

MRI phantoms more generally use aqueous paramagnetic solutions, gelatin, agar, silicone, agarose, or organically/paramagnetically doped gels. Dopants (nickel, manganese, gadolinium, aqueous oxygen) are used to adjust T1 and T2 to values appropriate for clinical pulse sequence timing.

A comprehensive systems test - assessing spatial resolution, statistical noise, count-rate performance, sensitivity, and image quality - should be performed at least annually.

Ultrasound Equipment QC

Ultrasound QC relies on tissue-mimicking phantoms and structured performance testing, complemented by daily informal assessment by the sonographer during routine clinical use. QC is essentially performed every day during routine scanning by the sonographer, who should recognise major problems; however, a formal periodic QC programme is required to detect problems before serious malfunctions occur.

Tissue-mimicking phantom material: attenuation coefficient of approximately $0.5$ to $0.7~\text{dB} \cdot \text{cm}^{-1} \cdot \text{MHz}^{-1}$ (the higher value provides a more challenging penetration test) and a speed of sound representative of soft tissue.

A standard general-purpose phantom comprises three modules:

Additional low-contrast sphere inserts assess contrast resolution.

Each transducer and transducer port must be tested individually. Additional system capabilities (harmonic imaging, elastography, colour flow Doppler) must be evaluated and documented to meet published specifications at acceptance. Transducer surface integrity should be checked for delamination or cracked face.

Testing frequency: establish empirically - perform tests frequently initially, review logbooks over an extended period, and with documented stability, reduce the testing rate as appropriate. This risk-informed approach allocates QC effort where instability is most likely.

Power output (acoustic output) of transducers can be assessed by weighing sound pressure with a force balance or measuring heating effect using a calorimeter; more sophisticated methods characterise the intensity or pressure distribution across the field.

Roles and Responsibilities in QA

The diagnostic medical physicist is responsible for testing imaging equipment initially, annually, and after repairs or modifications that may affect radiation doses to patients or image quality. This includes dose calibrator testing, gamma camera performance surveys, MRI system surveys, and CT dose verification.

Acceptance Testing vs. Routine QC vs. Annual Survey

Acceptance testing is the most comprehensive phase. For all modalities it establishes the baseline against which subsequent QC measurements are compared. Parameters outside manufacturer-specified tolerance at acceptance must be resolved before the equipment enters clinical use.

Documentation and Corrective Action

An effective QA programme requires:

1. Written programme document: goals, responsibilities, test procedures, equipment, tolerances, and corrective action thresholds
2. Logbooks: dated records of all QC measurements, the operator, and equipment status
3. Corrective action log: problem identified, investigation performed, action taken, verification of resolution, and date of return to service
4. Trend analysis: graphical display of QC metrics over time to detect gradual drift before action limits are breached
5. Clinical image quality feedback loop: radiologists must have a formal mechanism to report image quality concerns; in screening this is especially important as missed lesions can drive equipment re-evaluation

For mammography, the EQUIP framework mandates that the Lead Interpreting Radiologist review all required performance tests, document corrective actions, engage in regular image quality audits of technologist performance, and ensure mechanisms are in place for ongoing radiologist feedback on image quality - embedding clinical feedback within the technical QC cycle.

Summary: Key QC Parameters by Modality

A robust QA programme across all imaging modalities ensures that patients - both symptomatic and asymptomatic - receive examinations that are diagnostically adequate, dose-optimised, and performed on equipment proven to meet defined performance standards. For screening populations in particular, the stakes of equipment underperformance are amplified across large numbers of healthy individuals, making rigorous, well-documented, and regularly reviewed QC essential to programme integrity.