Maintaining Excellence in Medical Imaging Operations
Quality Assurance (QA) programs ensure consistent, high-quality medical imaging through systematic monitoring, testing, and improvement. A comprehensive QA program encompasses equipment performance, image quality, radiation safety, and clinical processes.
Regular testing ensures imaging equipment performs within specifications:
| Test | Frequency | Acceptance Criteria |
|---|---|---|
| Water CT Number | Daily | 0 ± 4 HU |
| CT Number Uniformity | Monthly | ±5 HU center to periphery |
| Noise (Standard Deviation) | Monthly | Within manufacturer specs |
| Spatial Resolution | Annual | Per stated specification |
| Low Contrast Resolution | Annual | Per stated specification |
| Radiation Dose (CTDI) | Annual | Within reference levels |
| Test | Frequency | Purpose |
|---|---|---|
| Center Frequency | Daily | Verify magnetic field stability |
| SNR | Weekly | Coil and system performance |
| Geometric Accuracy | Monthly | Spatial distortion within limits |
| Slice Thickness | Monthly | Accurate slice profiles |
| Image Uniformity | Monthly | Consistent signal across FOV |
| Ghosting | Annual | Artifact identification |
Phantoms are test objects with known properties used to evaluate equipment performance. Common phantoms include the ACR CT phantom (water, air, acrylic inserts), ACR MRI phantom (multiple test structures), and various modality-specific designs. Consistent phantom positioning and acquisition protocols are essential for meaningful trending.
Radiation dose optimization follows the ALARA principle—As Low As Reasonably Achievable while maintaining diagnostic quality:
| Metric | Modality | Description |
|---|---|---|
| CTDIvol | CT | Volume CT dose index (mGy) |
| DLP | CT | Dose-length product (mGy·cm) |
| SSDE | CT | Size-specific dose estimate |
| DAP/KAP | Fluoroscopy | Dose area product (Gy·cm²) |
| MGD | Mammography | Mean glandular dose (mGy) |
| Effective Dose | All | Whole-body risk estimate (mSv) |
DRLs are benchmark dose values representing typical doses for standard examinations. They are not limits but investigation levels—facilities should review protocols if doses consistently exceed DRLs. National DRLs (e.g., ACR DIR data) help identify optimization opportunities.
Medical displays require calibration for accurate image interpretation:
| Standard | Application | Key Requirements |
|---|---|---|
| DICOM GSDF | Grayscale calibration | Perceptually linear display |
| AAPM TG-18 | QC test patterns | Uniformity, resolution, noise |
| ACR Guidelines | Mammography displays | 5MP, calibrated luminance |
Display luminance decreases over time (typically 10-15% per year). Regular calibration and replacement of monitors approaching end-of-life is essential. Commercial display management software automates QC testing and tracking for enterprise deployments.
Accreditation demonstrates quality through external evaluation:
| Organization | Programs | Requirements |
|---|---|---|
| ACR | CT, MRI, US, Mammo, Nuclear | Phantom images, clinical images, survey |
| The Joint Commission | Hospital imaging services | Organization-wide quality standards |
| IAC | Echo, Vascular, Nuclear, CT, MRI | Modality-specific clinical standards |
Systematic incident management improves safety and quality:
Significant incidents require root cause analysis (RCA) to identify contributing factors and prevent recurrence. RCA examines system factors—not individual blame—including communication, training, equipment design, and workflow vulnerabilities.
Quality improvement programs drive ongoing enhancement:
| Category | Example Metrics | Target |
|---|---|---|
| Timeliness | Report turnaround time | <2 hours routine, <1 hour stat |
| Accuracy | Discrepancy rate | <3% major discrepancy |
| Safety | Contrast reaction rate | <0.5% |
| Patient Experience | Wait times, satisfaction | >90% satisfaction |
| Appropriateness | Protocol adherence | >95% appropriate use |
Successful QA programs: designate a qualified medical physicist; establish written policies and procedures; maintain detailed records; trend data over time; address failures promptly; engage staff at all levels; pursue accreditation; and foster a culture of continuous improvement.
WIA-Official/wia-standards-public/tree/main/medical-imaging — open standard initiative providing source code for simulator, spec, API, and ebook assets cited throughout this volume; serves as the canonical verification record for all primary-source citations made by the WIA standard committee in this chapter. Canonical ENUM tokens used in this volume include CT, MRI, PET, SPECT, MAMMOGRAPHY, ULTRASOUND, XRAY, DICOM_3_0, NIFTI, NRRD, JPEG_BASELINE, JPEG_2000_LOSSLESS, HEVC, DICOMWEB, QIDO_RS, STOW_RS, WADO_RS, PACS, VNA, RIS, HIS, FDA_510K, CE_MDR, MFDS_CLASS_2, HIPAA, GDPR, DICOM_CONFORMANCE, RUL, RML, RLL, LUL, LLL, MEDIASTINUM, HL7_FHIR_IMAGING, IHE_RAD, NEMA_MITA, DICOM_SR, RADLEX, U_NET, V_NET, RESNET, VIT, TRANSFORMER, CNN, MONAI.This section aligns Chapter 7 quality assurance content with simulator panels 2 (Protocol) and 4 (Test) at the token level. The certification and regulatory ENUMs DICOM_CONFORMANCE, FDA_510K, CE_MDR, MFDS_CLASS_2, HIPAA, and GDPR appear throughout the book in UNDERSCORE notation, with test items, acceptance criteria, and documentation requirements verified by the Standards Revision Committee. Standard ENUMs DICOM_3_0, DICOMWEB, IHE_RAD, and NEMA_MITA apply to PACS, VNA, and RIS interoperability testing.
| ENUM Token | Regulation | Authority | Sections |
|---|---|---|---|
FDA_510K | 510(k) substantial equivalence pathway | US FDA | §7.4, §7.5 |
CE_MDR | EU Medical Device Regulation 2017/745 | EU Notified Body | §7.4, §7.6 |
MFDS_CLASS_2 | MFDS Class 2 medical device | Korea MFDS | §7.4 |
HIPAA | US Health Insurance Portability and Accountability Act | HHS OCR | §7.6 |
GDPR | EU General Data Protection Regulation | EU DPAs | §7.6 |
DICOM_CONFORMANCE | DICOM Conformance Statement | NEMA MITA | §7.1, §7.7 |
Simulator panel 4 (Test) simulates QA thresholds across modalities. CT computes CTDIvol, DLP, and SSDE thresholds by patient size and exam type; MRI evaluates SNR, CNR, geometric accuracy, and coil uniformity; MAMMOGRAPHY measures ACR Phantom score and average glandular dose (AGD); ULTRASOUND verifies beam profile, axial resolution, and lateral resolution; PET and SPECT assess spatial resolution, sensitivity, and scatter fraction per NEMA NU-2; and DXA evaluates BMD reproducibility (CV < 1%). Pulmonary lobe ENUMs RUL, RML, RLL, LUL, LLL, MEDIASTINUM, and HILUM serve as standard labels when verifying chest imaging segmentation outputs.
ISO 13485:2016 governs quality management systems across the entire medical device lifecycle: manufacturing, distribution, and service. Imaging device manufacturers, Software as a Medical Device (SaMD) developers, and PACS vendors all require ISO 13485 certification. Core requirements span (1) quality policy and objectives, (2) management responsibility, (3) resource management, (4) product realization (design, procurement, production, service), and (5) measurement, analysis, and improvement. The standard explicitly addresses risk-based decision making throughout the lifecycle. ACR Practice Parameters supplement modality-specific QC frequencies and acceptance criteria.
ISO 14971:2019 specifies risk management processes for medical devices: Hazard Analysis, Risk Evaluation, Risk Control, Residual Risk evaluation, and Post-Production Information form the five-stage lifecycle, governed by the ALARP (As Low As Reasonably Practicable) principle. Representative hazards for imaging devices include (a) radiation overexposure, (b) magnetic field exposure (MRI), (c) contrast agent reactions, (d) image processing errors (artifacts, reconstruction failure), and (e) cybersecurity breaches. AAPM Report No. 96 is the standard reference document for CT dose risk management.
IEC 62366-1:2015+A1:2020 specifies the usability engineering process for medical devices. To prevent patient harm from use errors, the standard mandates specification of intended use, users, and use environment, plus validation of the user interface design. PACS, viewers, and AI analysis tools require both Formative Evaluation (during design) and Summative Evaluation (after design completion). The standard interfaces tightly with ISO 14971 risk management for use-related hazards. AAPM Task Group 18 (2005) provides standard test procedures for medical displays.
The DICOM Conformance Statement documents which SOP Classes, transfer syntaxes, and parameters a medical device supports. NEMA MITA provides the standard template, and imaging devices, PACS, viewers, and AI analysis tools must all publish one. The annual IHE Connectathon brings multinational medical device manufacturers together for interoperability testing; products that pass receive IHE Profile compliance labeling, valuable for procurement decisions. DICOM PS3.14 "Grayscale Standard Display Function (GSDF)" is the core standard for display calibration.
Corrective and Preventive Action (CAPA) is a core requirement of ISO 13485 and FDA QSR (21 CFR 820). When nonconformance is identified, the four-step process applies: (1) root cause analysis, (2) corrective action, (3) preventive action, and (4) effectiveness verification. Retention of CAPA records typically spans the device lifetime plus 2 years; for imaging devices this commonly translates to 10-15 years. FDA 21 CFR Part 11 adds validation requirements for electronic records and electronic signatures.
ACR (American College of Radiology) Practice Parameters define clinical imaging standards by modality and anatomical region. ACR Accreditation Programs evaluate facility qualifications for CT, MRI, mammography, PET, nuclear medicine, ultrasound, and breast MRI through phantom imaging, clinical imaging, and QA record review. Equivalent national programs exist in many countries; in Korea, the Korea Foundation of Radiation Medicine (KFRM) and the Korean Society of Radiology (KSR) operate parallel accreditation frameworks. The Joint Commission "Diagnostic Imaging Standards" (2024) and the Intersocietal Accreditation Commission (IAC) provide complementary accreditation pathways.
connectathon.ihe.net, 2024 ────.