Discover the 17 most common CT artifacts that compromise diagnostic accuracy, from beam hardening streaks to quantum mottle noise, with proven remedies and real clinical examples for radiology teams.

17 Types of CT Artifacts: Causes, Remedies & Clinical Examples Every Radiologist Must Know

⏱️ 22 min read 📁 Diagnostic Imaging ✓ Medically Reviewed

At a glance

• CT artifacts are systematic image distortions that can mimic pathology, obscure findings, or lead to misdiagnosis
• Artifacts are broadly classified as patient-related (motion, metal), physics-related (beam hardening, scatter), and equipment-related (ring, calibration errors)
• Modern metal artifact reduction (MAR) algorithms and iterative reconstruction can significantly improve image quality
• Understanding artifact mechanisms enables radiographers to adjust protocols proactively rather than repeating scans
• This guide covers 17 distinct artifact types with visual examples and evidence-based remedies for each

Introduction to CT artifacts

Computed tomography has revolutionised diagnostic imaging since Hounsfield’s Nobel Prize-winning invention in 1979, yet every radiographer and radiologist encounters images compromised by artifacts that threaten diagnostic confidence. CT artifacts are defined as any systematic discrepancy between the reconstructed CT numbers and the true linear attenuation coefficients of the tissues being imaged. These distortions can mimic pathology, obscure critical findings, or create false-positive interpretations that lead to unnecessary interventions.

The clinical impact of CT artifacts extends beyond mere image aesthetics. In emergency settings, a beam hardening streak across the posterior fossa can obscure a cerebellar haemorrhage. In oncology, metal streaking from hip prostheses can render pelvic lymph nodes unassessable. In paediatrics, motion artifacts from an uncooperative child may necessitate sedation and repeat radiation exposure. Understanding artifact mechanisms is therefore not merely academic—it is essential for patient safety and diagnostic accuracy.

This comprehensive guide examines 17 distinct CT artifact categories, organised by their underlying physical mechanism. For each artifact type, we present: the physics of origin, characteristic imaging appearance, clinical scenarios where it commonly occurs, evidence-based remedies, and illustrative examples. The goal is to equip radiographers, radiologists, and hospital administrators with practical knowledge to optimise CT protocols and minimise diagnostic errors.

1. Beam hardening artifacts

Physics and mechanism

Beam hardening occurs because X-ray tubes produce a polychromatic spectrum of photon energies, typically ranging from approximately 20 keV to 140 keV for a 120 kVp tube voltage. As the beam traverses dense tissues—particularly bone, iodinated contrast, or metal—lower-energy photons are preferentially absorbed, leaving a “harder” (higher average energy) beam. The reconstruction algorithm assumes a monochromatic beam, so the measured attenuation does not match the expected values.

Two primary manifestations result from beam hardening: cupping artifacts and streaking artifacts. Cupping produces a gradual darkening toward the centre of a uniformly dense object, while streaking creates dark bands between two dense objects or bright streaks radiating from a single dense focus. The Hounsfield unit (HU) values in affected regions become unreliable, potentially causing misinterpretation of tissue density.

Clinical examples

Beam hardening is most problematic in the posterior fossa, where the dense petrous temporal bones create severe streaking that obscures the brainstem and cerebellum. In thoracic CT, the shoulders and clavicles generate streaks across the lung apices. In abdominal CT, contrast-filled vessels and calcified atherosclerotic plaques produce cupping that can simulate low-attenuation lesions.

Remedies and solutions

• Software correction: All modern CT scanners apply built-in beam hardening correction (BHC) algorithms. Ensure these are enabled in the protocol.
• Dual-energy CT: Virtual monoenergetic images at higher keV (e.g., 140 keV) virtually eliminate beam hardening by synthesising monochromatic data.
• Iterative reconstruction: Advanced algorithms like SAFIRE, ASiR-V, and ADMIRE model polychromatic physics more accurately than filtered back projection.
• Positioning: Avoid placing dense anatomy symmetrically in the beam path when possible.
• Post-processing: Apply edge-enhancing kernels and adjust window settings to minimise visual impact.

2. Metal streaking artifacts

Physics and mechanism

Metal streaking represents the most severe form of beam hardening, occurring when high-atomic-number materials (titanium, cobalt-chrome, stainless steel, dental amalgam) completely attenuate the X-ray beam. The resulting photon starvation at detector elements behind the metal causes the reconstruction algorithm to generate extreme streaking radiating from the metal object. Additionally, scatter radiation and partial volume effects compound the problem.

The severity depends on the metal’s composition: stainless steel produces the worst artifacts due to its high density and atomic number, while titanium causes less severe but still significant streaking. Dental fillings, surgical clips, hip prostheses, spinal fixation hardware, and aneurysm coils are common culprits.

Clinical examples

In head and neck CT, dental amalgam fillings create streaks that obscure oral cavity tumours and mandibular fractures. In orthopaedic CT, bilateral hip replacements can render the entire pelvis non-diagnostic. In neurovascular CT, aneurysm clips and coils produce artifacts that limit follow-up assessment.

Remedies and solutions

• Metal Artifact Reduction (MAR) algorithms: Commercial solutions include O-MAR (Philips), iMAR (Siemens), SEMAR (Canon), and Smart MAR (GE). These replace metal-affected projections with interpolated data.
• Dual-energy CT with virtual monoenergetic images: High-keV images reduce metal-induced streaking by 60-80%.
• Iterative Metal Artifact Reduction (iMAR): Combines projection completion with iterative reconstruction for superior results.
• Scan parameter optimisation: Use highest possible kVp (140 kVp), increased mAs, and smaller pitch to improve photon statistics.
• Patient positioning: Angle gantry to avoid direct beam path through metal when anatomy permits.
• Pre-scan screening: Identify metal before scanning to select appropriate protocol.

3. Ring artifacts

Physics and mechanism

Ring artifacts arise from detector element miscalibration or failure. In third-generation CT scanners with rotate-rotate geometry, each detector element traces a circular path during acquisition. If one element consistently over-reads or under-reads, the error projects as a ring at a specific radius from the isocentre. Causes include detector afterglow, dead pixels, scintillator degradation, or reference channel malfunction.

Ring artifacts are particularly problematic in flat-panel detector CT and Cone Beam CT (CBCT) used in interventional radiology and radiation therapy, where detector calibration is more challenging than in conventional multidetector CT.

Clinical examples

A single ring artifact through the posterior fossa can be mistaken for a subarachnoid haemorrhage or calcified lesion. Multiple concentric rings create a “bulls-eye” pattern that severely degrades image quality throughout the slice.

Remedies and solutions

• Detector calibration: Perform air calibration and water phantom QC daily or per manufacturer schedule.
• Detector replacement: Persistent rings indicate hardware failure requiring service engineer intervention.
• Software ring correction: Most scanners include ring artifact suppression algorithms that interpolate around faulty channels.
• Dual-energy CT: Material decomposition can reduce ring visibility in some implementations.
• Preventive maintenance: Monitor detector temperature stability and humidity control in scanner room.

4. Motion artifacts

Physics and mechanism

Motion artifacts occur when the patient moves during data acquisition, causing projection inconsistencies that the reconstruction algorithm cannot reconcile. The resulting images show blurring, ghosting, or streaking in the direction of motion. Voluntary motion (patient discomfort, anxiety) and involuntary motion (cardiac pulsation, respiration, peristalsis) both contribute.

The severity depends on motion speed relative to scan time. On conventional CT, a full rotation takes 0.5-1.0 seconds, so even slight head movement degrades image quality. Cardiac motion causes blurring of coronary arteries and valve leaflets. Respiratory motion creates artifacts at the lung bases and diaphragm.

Clinical examples

Motion in head CT creates bizarre streaking patterns that can be mistaken for subarachnoid blood or fracture lines. In chest CT, respiratory motion causes pseudo-ground-glass opacities. In abdominal CT, peristalsis creates streaking that mimics bowel perforation.

Remedies and solutions

• Patient coaching: Clear breathing instructions and immobilisation are the most effective preventive measures.
• Immobilisation devices: Head straps, foam cushions, and vacuum mattresses reduce voluntary motion.
• Faster scan times: Use highest gantry rotation speed available (sub-0.5 second on modern scanners).
• Prospective/retrospective gating: For cardiac CT, ECG-gating synchronises acquisition to cardiac phase.
• Motion correction algorithms: Some vendors offer data-driven motion correction that re-aligns projections.
• Sedation: For paediatric or claustrophobic patients, consider conscious sedation protocols.

5. Partial volume averaging

Physics and mechanism

Partial volume averaging occurs when a single voxel contains multiple tissue types with different attenuation coefficients. The reconstructed CT number represents the weighted average of all tissues within the voxel, potentially obscuring small lesions or creating false interfaces. This is fundamentally a sampling limitation related to voxel size.

The effect is most pronounced with small structures (sub-centimetre nodules, thin bone trabeculae) and at tissue interfaces (lung-soft tissue, bone-marrow). A 5 mm lung nodule in a 5 mm slice may be completely averaged with adjacent aerated lung, reducing its apparent density and causing underestimation of enhancement.

Clinical examples

In lung cancer screening, small subsolid nodules can appear purely ground-glass or solid depending on slice thickness. In renal CT, small cysts may show pseudo-enhancement due to averaging with enhancing parenchyma. In bone CT, trabecular detail is lost with thick slices.

Remedies and solutions

• Thin slice acquisition: Reconstruct at ≤1.25 mm for small lesion detection; review images at narrow slice thickness.
• Overlapping reconstruction: Use 50% overlap to improve z-axis sampling.
• Sharp kernels: Edge-enhancing algorithms reduce partial volume blurring.
• Dual-energy CT: Material-specific images can separate mixed materials within a voxel.
• Volume rendering: 3D visualisation helps appreciate true lesion morphology.

6. Quantum mottle (noise)

Physics and mechanism

Quantum mottle is the statistical noise arising from the finite number of X-ray photons detected. It follows Poisson statistics: the signal-to-noise ratio (SNR) equals the square root of the number of detected photons. Low-dose protocols, high-attenuation anatomy (obese patients), and thin slices all increase quantum mottle.

Quantum mottle appears as a grainy texture superimposed on the image, reducing low-contrast detectability. While it is technically a noise rather than a true artifact, its impact on diagnostic quality warrants inclusion in any comprehensive CT quality discussion.

Clinical examples

In low-dose lung screening CT, quantum mottle can obscure subtle ground-glass opacities. In CT colonography, noise may obscure small polyps. In paediatric CT, where dose reduction is paramount, balancing noise and diagnostic quality is critical.

Remedies and solutions

• Dose modulation: Use automatic tube current modulation (Care Dose 4D, Smart mA, etc.) to adapt mA to patient anatomy.
• Iterative reconstruction: Model-based iterative reconstruction (MBIR) can reduce dose by 50-80% while maintaining noise levels.
• Thicker slices: Increasing slice thickness from 1 mm to 3 mm reduces noise by √3.
• Higher kVp: In large patients, 140 kVp improves penetration and photon statistics.
• Smoothing kernels: Soft-tissue kernels reduce noise at the expense of spatial resolution.

7. Scatter radiation artifacts

Physics and mechanism

Scatter radiation occurs when X-ray photons undergo Compton interactions within the patient, deflecting from their original path and reaching detector elements at incorrect angles. Scattered photons carry false positional information, causing cupping, streaking, and reduced contrast. The scatter-to-primary ratio increases with patient size and field-of-view.

In CBCT and wide-detector CT, scatter is particularly problematic because anti-scatter grids are less effective or absent. In dual-energy CT, scatter can corrupt material decomposition accuracy.

Clinical examples

Scatter from the shoulders in thoracic CT causes dark bands across the lung apices. In obese patients, scatter contributes to global image degradation and CT number inaccuracy.

Remedies and solutions

• Anti-scatter grids: Ensure grid is engaged for body CT; remove for head CT to improve dose efficiency.
• Software scatter correction: Monte Carlo-based scatter correction is available on some advanced systems.
• Collimation: Tight pre-patient collimation reduces scatter volume.
• Air gap: Increasing source-to-detector distance reduces scatter acceptance (used in CBCT).
• Dual-energy scatter correction: Energy-resolved detection can separate scatter from primary photons.

8. Windmill artifacts

Physics and mechanism

Windmill artifacts are specific to helical CT with multi-slice detectors. They appear as periodic streaking radiating from high-contrast edges, resembling a windmill or propeller pattern. The artifact arises from the helical interpolation process, where data from different z-positions and angles are combined. When the pitch is high and the object has sharp edges, the interpolation creates periodic inconsistencies.

Clinical examples

Windmill artifacts are most visible around metallic objects, dense bone edges, and contrast-filled vessels. They can simulate dissection flaps in aortic CT angiography or obscure stent lumens.

Remedies and solutions

• Reduce pitch: Lower pitch values (0.6-1.0) reduce windmill severity at the cost of increased dose.
• Non-helical scanning: For small volumes with metal, axial step-and-shoot acquisition eliminates helical interpolation.
• Adaptive collimation: Some scanners adjust collimator position to reduce windmill patterns.
• Iterative reconstruction: Advanced algorithms better handle helical interpolation errors.

9. Cone beam artifacts

Physics and mechanism

Cone beam artifacts occur in CT systems with wide-area detectors where the X-ray beam forms a cone rather than a fan. The Feldkamp-Davis-Kress (FDK) algorithm assumes parallel geometry, which becomes increasingly inaccurate at the cone edges. This causes geometric distortion, streaking, and intensity drop-off at the periphery of the field-of-view.

These artifacts are inherent to Cone Beam CT (CBCT) used in dental imaging, interventional radiology, and radiation therapy planning. The wider the cone angle, the more severe the artifacts.

Clinical examples

In dental CBCT, streaking from metallic restorations is amplified by cone beam geometry. In interventional CBCT, peripheral structures show geometric distortion. In radiation therapy, CBCT-based dose calculations may be inaccurate at field edges.

Remedies and solutions

• Field-of-view limitation: Restrict scanning to the region of interest.
• Multiple acquisitions: Stitch narrow FOV scans for full coverage.
• Advanced reconstruction: Katsevich-type exact reconstruction algorithms reduce cone beam errors.
• Detector calibration: Geometric calibration is critical for CBCT accuracy.

10. Truncation artifacts

Physics and mechanism

Truncation artifacts occur when the patient extends beyond the scan field-of-view (SFOV), typically in large patients or when arms are not raised. The reconstruction algorithm assumes zero attenuation outside the measured region, causing bright streaks at the truncation boundary and global CT number inaccuracy.

Clinical examples

In bariatric CT, lateral trunk soft tissues are truncated, causing bright streaks across the abdomen. In shoulder CT with arms at sides, the arms may be truncated, affecting thoracic image quality.

Remedies and solutions

• Patient positioning: Raise arms above head for thoracic and abdominal CT.
• Larger SFOV: Use extended field-of-view modes (70 cm) when available.
• Software correction: Water cylinder extrapolation algorithms estimate truncated attenuation.
• Dual acquisition: Scan large patients in two overlapping acquisitions.

11. Stair-step artifacts

Physics and mechanism

Stair-step artifacts appear in multiplanar reconstructions (MPR) and 3D volume renderings when adjacent slices have different effective z-positions due to table movement between rotations. The artifact manifests as a stepped appearance along oblique surfaces, particularly visible on curved structures like ribs and vessels.

Clinical examples

Stair-step artifacts on CT angiography MPRs can simulate vessel wall irregularity or dissection. In cardiac CT, they affect coronary artery assessment.

Remedies and solutions

• Overlapping reconstruction: 50% overlap minimises stair-stepping.
• Thinner slices: Smaller z-axis sampling interval reduces step visibility.
• Smoothing filters: Apply to MPR before 3D rendering.
• Retrospective gating: For cardiac CT, ensures consistent phase.

12. Off-focus radiation

Physics and mechanism

Off-focus radiation consists of X-rays produced outside the focal spot, typically from electron backscatter onto the anode. These extra-focal photons create a halo around dense objects and reduce edge sharpness. The effect is more pronounced with larger focal spots and higher tube currents.

Clinical examples

Off-focus radiation causes edge blurring around bone and contrast-filled vessels, reducing spatial resolution in high-detail examinations like temporal bone CT and CT urography.

Remedies and solutions

• Small focal spot: Use when spatial resolution is critical.
• Pre-patient collimation: Tight collimation reduces off-focus acceptance.
• Post-patient collimation: Detector-side collimation further restricts acceptance angle.

13. Helical interpolation artifacts

Physics and mechanism

Beyond windmill artifacts, helical interpolation can cause z-axis blurring and slice sensitivity profile broadening at high pitches. The 360° linear interpolation (360LI) or 180° linear interpolation (180LI) algorithms combine data from different z-positions, causing effective slice thickness to exceed nominal thickness.

Clinical examples

Small pulmonary nodules may appear artificially enlarged and less dense due to z-axis blurring. In pancreatic CT, subtle ductal dilatation may be obscured.

Remedies and solutions

• Lower pitch: Pitch ≤1.0 minimises interpolation errors.
• Non-helical acquisition: For small, well-defined volumes.
• Sharp kernels: Compensate for z-axis blurring.

14. Cupping artifacts

Physics and mechanism

Cupping is a specific manifestation of beam hardening where a uniformly dense cylindrical object (like the head or abdomen) appears darker in the centre than at the periphery. The “cup” shape of the HU profile across the diameter gives the artifact its name. It occurs because the beam hardens more as it passes through the centre (longer path length) than at the edges.

Clinical examples

In head CT, cupping can simulate hypodense white matter disease or oedema. In phantom QC, cupping indicates inadequate beam hardening correction.

Remedies and solutions

• Beam hardening correction: Ensure BHC is enabled and calibrated.
• Dual-energy CT: Virtual monoenergetic images eliminate cupping.
• Water phantom QC: Verify cupping is within manufacturer tolerance.

15. Out-of-field artifacts

Physics and mechanism

Out-of-field artifacts occur when anatomy outside the reconstructed field-of-view still attenuates the X-ray beam. The reconstruction algorithm has no data for these regions, causing bright streaks at the FOV boundary and CT number shifts within the image.

Clinical examples

In head CT with small FOV, the skull base may extend beyond the reconstruction circle, causing artifacts. In dental CBCT, the tongue and soft palate outside the dental FOV create streaks.

Remedies and solutions

• Larger reconstruction FOV: Always reconstruct to include all attenuating anatomy.
• Patient positioning: Centre anatomy within the FOV.
• Extended reconstruction: Some scanners offer extended FOV reconstruction.

16. Calibration errors

Physics and mechanism

Calibration errors arise from incorrect mapping between detector signal and Hounsfield units. Causes include reference detector drift, kVp instability, bowtie filter misalignment, and detector temperature fluctuations. These cause global or regional CT number inaccuracy.

Clinical examples

A calibration error causing +50 HU shift could make a simple renal cyst appear enhancing, triggering unnecessary workup. In lung nodule tracking, HU changes may be artifactual rather than representing true growth.

Remedies and solutions

• Daily air calibration: Mandatory before first patient.
• Water phantom QC: Weekly HU accuracy verification.
• Temperature monitoring: Maintain scanner room within specification.
• Service intervention: For persistent calibration drift.

17. Photon-counting detector artifacts

Physics and mechanism

Photon-counting detectors (PCDs) represent the next generation of CT technology, replacing energy-integrating detectors with semiconductor materials (cadmium telluride, cadmium zinc telluride) that count individual photons and classify them by energy. While PCDs dramatically reduce many conventional artifacts—beam hardening, electronic noise, and scatter—they introduce new artifact types that radiologists must recognise.

Charge sharing occurs when a single photon interaction deposits energy in multiple detector pixels, causing the photon to be counted at an incorrect energy or position. Pulse pile-up happens when two photons arrive within the detector’s dead time, creating a single higher-energy count. K-escape occurs when characteristic X-rays from the detector material escape before full energy deposition, creating low-energy counts. Compton scatter within the detector causes photons to be counted at incorrect energies.

Clinical examples

In PCD-based dual-energy CT, charge sharing can corrupt material decomposition accuracy, causing iodine to be misclassified as calcium or vice versa. In high-flux imaging (obese patients, high kVp), pulse pile-up reduces the effective count rate and degrades image quality. In small FOV imaging (dental, extremity), K-escape creates edge artifacts at the field periphery.

Remedies and solutions

• Anti-charge sharing grids: Physical septa between pixels reduce charge sharing.
• Software correction: Real-time algorithms detect and correct pile-up and charge sharing events.
• Energy threshold optimisation: Careful selection of energy bin thresholds minimises K-escape effects.
• Flux management: Automatic tube current modulation prevents detector saturation in high-flux scenarios.
• Regular calibration: PCDs require more frequent energy calibration than conventional detectors.

Advanced artifact reduction techniques

Artificial intelligence in CT artifact reduction

Deep learning has emerged as a powerful tool for CT artifact reduction. Convolutional neural networks (CNNs) can be trained to map artifact-corrupted images to artifact-free ground truth, learning complex non-linear relationships that traditional algorithms cannot capture. Generative adversarial networks (GANs) have shown particular promise for metal artifact reduction, with some studies demonstrating image quality improvements equivalent to a 50% reduction in metal-induced streaking.

Commercial implementations include TrueFidelity (GE Healthcare), which uses a deep neural network trained on thousands of clinical cases to produce noise-free images from low-dose acquisitions. PureVision (Siemens) employs a similar approach with manufacturer-specific training. These systems are particularly valuable in paediatric CT, where dose reduction is paramount, and in obese patient imaging, where quantum mottle traditionally limits diagnostic quality.

For motion artifact correction, deep learning models analyse projection data to detect and compensate for patient movement without the need for external tracking devices. Data-driven motion correction algorithms can reduce motion artifacts by 60-80% in head CT, potentially eliminating the need for repeat scans in uncooperative patients.

Dual-energy CT and artifact management

Dual-energy CT (DECT) is arguably the most significant advance in CT artifact reduction since iterative reconstruction. By acquiring data at two different kVp settings (or using dual-layer detectors), DECT enables:

• Virtual monoenergetic images: Synthesise images at any keV between 40 and 200 keV. High-keV images (140-200 keV) virtually eliminate beam hardening and metal streaking.
• Material decomposition: Separate materials by their energy-dependent attenuation properties, enabling iodine mapping, calcium quantification, and uric acid detection.
• Virtual non-contrast images: Subtract iodine from contrast-enhanced scans to create unenhanced-equivalent images, reducing the need for pre-contrast acquisitions.
• Effective atomic number mapping: Identify unknown materials by their effective Z value.

Clinical applications of DECT for artifact reduction include: metal artifact reduction in orthopaedic and dental imaging (60-80% streak reduction), beam hardening correction in posterior fossa and skull base imaging, and contrast optimisation in patients with impaired renal function who cannot receive full contrast doses.

Iterative reconstruction evolution

Iterative reconstruction has evolved through three generations: hybrid iterative reconstruction (first generation, e.g., ASiR, iDose, SAFIRE), which combines filtered back projection with statistical noise modelling; model-based iterative reconstruction (second generation, e.g., Veo, FIRST, ADMIRE), which models the complete imaging chain including X-ray spectrum, detector response, and object physics; and deep learning reconstruction (third generation, e.g., TrueFidelity, AiCE, Photon), which uses neural networks trained on clinical data.

Each generation offers progressively better artifact handling. Model-based methods reduce beam hardening by modelling the polychromatic spectrum. Deep learning methods can learn to recognise and suppress artifacts from any source, including those not explicitly modelled in the physics. The trade-off is computational cost: deep learning reconstruction requires GPU acceleration and can add 30-60 seconds to reconstruction time.

Artifacts that mimic pathology: a differential diagnosis

Perhaps the greatest clinical danger of CT artifacts is their ability to mimic genuine pathology, leading to misdiagnosis, unnecessary procedures, and patient harm. The following table summarises the most common artifact-mimic pairs and how to distinguish them:

Artifact	Mimics	Differentiating Features
Beam hardening streak	Subarachnoid haemorrhage, fracture line	Connects two dense objects; changes with window; absent on dual-energy
Metal streaking	Tumour, abscess	Radiates from known metal; improves with MAR; absent on opposite side
Motion blur	Ground-glass opacity, infiltrate	Directional streaking; absent on repeat scan; correlates with patient movement
Partial volume	Ground-glass nodule, cyst	Changes with slice thickness; disappears on thin slices; no true lesion margins
Quantum mottle	Ground-glass opacity, interstitial disease	Random pattern; improves with thicker slices or higher dose; no true structure
Ring artifact	Calcified lesion, haemorrhage	Perfect circle centred on isocentre; appears on multiple slices; phantom confirms
Windmill	Dissection flap, stenosis	Periodic pattern; changes with pitch; absent on non-helical acquisition
Cupping	Hypodense lesion, oedema	Gradual centre-to-periphery gradient; present on uniform phantoms; corrects with BHC

CT quality assurance for artifact prevention

A comprehensive QA programme is essential for maintaining artifact-free imaging. The following schedule represents best practice based on ACR and ICRP guidelines:

Daily checks

• Air calibration: Mandatory before first patient. Corrects for detector drift and temperature changes.
• Water phantom scan: Verify CT number accuracy (0 ± 5 HU for water), noise level, and uniformity.
• Visual inspection: Check for ring artifacts, noise patterns, and image uniformity on the first patient scan.

Weekly checks

• Spatial resolution: Measure MTF using line pair phantoms or wire phantoms.
• Low-contrast detectability: Evaluate using ACR accreditation phantom or equivalent.
• Dose verification: Confirm CTDIvol matches expected values for standard protocols.

Monthly checks

• Comprehensive phantom analysis: Evaluate all image quality parameters using a multi-purpose phantom.
• Alignment verification: Check laser alignment, table position accuracy, and gantry tilt.
• Contrast resolution: Verify ability to detect low-contrast objects at specified dose levels.

Annual checks

• Full system calibration: Manufacturer service visit for comprehensive hardware verification.
• Radiation output verification: Independent measurement of tube output and beam quality.
• Software updates: Evaluate and install vendor software updates that may include new artifact correction algorithms.

Artifact considerations in special populations

Paediatric CT

Motion artifacts dominate paediatric CT, with up to 25% of examinations requiring sedation in children under 5 years. Protocol optimisation strategies include: fast scan times (sub-0.5 second rotation), immobilisation devices (vacuum mattresses, foam cushions), and child-friendly environments (decoration, music, parental presence). Dose reduction is paramount: use size-specific protocols with automatic tube current modulation, iterative reconstruction to maintain quality at low dose, and shielding for radiosensitive organs when appropriate.

Paediatric-specific artifacts include: breathing artifacts in infants who cannot hold their breath, contrast timing errors due to variable cardiac output, and motion from anxiety in older children. Parental coaching, distraction techniques, and when necessary, conscious sedation protocols are essential.

Obese patient CT

Obese patients (BMI >30) present unique artifact challenges. Photon starvation in thick body regions increases quantum mottle and reduces contrast resolution. Scatter radiation increases with patient size, causing cupping and streaking. Truncation artifacts are common when patients exceed the SFOV. Beam hardening is more severe due to longer path lengths.

Protocol adaptations for obese patients include: 140 kVp for improved penetration, increased mAs (often 2-4× standard), larger SFOV (70 cm), thicker slices to improve SNR, and iterative reconstruction to maintain quality at high dose. Dual-energy CT with high-keV images can improve image quality in extremely obese patients.

Emergency CT

In the emergency setting, scan time minimisation is paramount, but artifact management cannot be neglected. Trauma patients often have metallic fragments (bullets, shrapnel) that create severe streaking. Intubated patients have endotracheal tubes and monitoring lines that cause artifacts. Unconscious patients cannot follow breathing instructions, increasing motion artifacts.

Emergency-specific strategies include: pre-scan screening for metal, MAR protocols for known implants, fastest available scan modes, and retrospective gating for cardiac motion when ECG is available. In polytrauma, accept some artifact in non-critical regions to minimise scan time and radiation dose.

Oncology CT

In oncology imaging, artifact management is critical for response assessment and follow-up. Metal artifacts from surgical clips, ports, and stents can obscure tumour margins. Partial volume effects cause underestimation of small lesion size. Contrast timing must be consistent between baseline and follow-up scans to ensure comparable enhancement patterns.

Oncology-specific protocols should include: standardised scan parameters for serial comparisons, thin slices for small lesion assessment, MAR algorithms for patients with surgical hardware, and dual-energy CT for iodine quantification in treatment response assessment.

Future directions in CT artifact management

The next decade promises transformative advances in CT artifact reduction. Photon-counting detectors will virtually eliminate beam hardening and electronic noise while enabling ultra-high-resolution imaging. AI-based reconstruction will learn to recognise and suppress artifacts from any source without explicit physics modelling. Motion-free cardiac CT will become routine with sub-millisecond scan times and AI-based motion correction.

Photon-counting CT represents the most significant hardware advance. By counting individual photons and classifying them by energy, PCDs eliminate the energy-integrating step that causes beam hardening. They also reduce electronic noise to negligible levels, enabling high-resolution imaging at ultra-low doses. The first clinical PCD-CT systems are now FDA-approved and entering clinical use.

Deep learning will transform artifact correction from a physics-based to a data-driven approach. Neural networks trained on millions of clinical cases will recognise artifact patterns and apply optimal corrections in real-time. Federated learning approaches will enable algorithms to improve continuously across multiple institutions without sharing patient data.

For radiology departments, the key to staying ahead of artifact challenges is continuous education, investment in QA infrastructure, and close collaboration between radiologists, radiographers, physicists, and vendor engineers. The radiologist who understands artifact physics will always deliver more reliable diagnoses than one who relies solely on image interpretation.

Summary of CT artifacts and remedies

Artifact	Primary Cause	Key Remedy	Clinical Priority
Beam hardening	Polychromatic spectrum	Dual-energy CT, BHC	High
Metal streaking	Photon starvation	MAR algorithms	Critical
Ring artifact	Detector miscalibration	Detector calibration/service	High
Motion	Patient movement	Faster scans, immobilisation	High
Partial volume	Voxel size limitation	Thin slices, overlap	Medium
Quantum mottle	Photon statistics	Iterative reconstruction	Medium
Scatter	Compton interactions	Anti-scatter grid	Medium
Windmill	Helical interpolation	Lower pitch	Medium
Cone beam	Wide detector geometry	Advanced reconstruction	Medium
Truncation	Patient exceeds SFOV	Positioning, larger FOV	High
Stair-step	Table movement	Overlapping reconstruction	Low
Off-focus	Extra-focal radiation	Small focal spot	Low
Cupping	Differential hardening	BHC, dual-energy	Medium
Calibration error	Detector drift	Daily QC	Critical

Further reading

1. SATLine: Advanced CT Quality Assurance Protocols — SATMED Health
2. SATDrape: Patient Positioning Systems for Artifact Reduction — SATMED Health
3. SATPro: Dual-Energy CT Clinical Applications — SATMED Health
4. SATSurgical: Metal Artifact Reduction in Orthopaedic Imaging — SATMED Health
5. SATJect: Contrast Protocol Optimisation for CT Angiography — SATMED Health

Conclusion

CT artifacts represent an inevitable consequence of the physical and technical limitations of computed tomography, yet their impact on diagnostic accuracy can be dramatically reduced through systematic understanding and proactive management. This guide has examined 17 distinct artifact categories, from the ubiquitous beam hardening streaks that plague posterior fossa imaging to the subtle calibration errors that can shift global CT numbers and mislead interpretation.

The key framework for artifact management follows three principles: prevention through optimised protocols and patient preparation, recognition through radiologist and radiographer education, and correction through advanced algorithms and hardware solutions. Modern iterative reconstruction, metal artifact reduction algorithms, and dual-energy CT have transformed what was once an insurmountable problem into a manageable challenge.

For hospital administrators, investing in comprehensive QC programmes and continuing education for radiography staff yields measurable returns in reduced repeat scan rates, improved diagnostic confidence, and enhanced patient safety. For radiologists, maintaining awareness of artifact appearances prevents the costly errors of false-positive diagnosis and missed pathology. For radiographers, understanding the physics behind each artifact empowers protocol optimisation at the scanner console.

As CT technology continues to evolve—with photon-counting detectors, AI-based reconstruction, and ever-faster acquisition—the artifact landscape will shift. Yet the fundamental physics of X-ray attenuation, photon statistics, and sampling theory remain unchanged. Mastery of these principles ensures that radiology professionals can adapt to new technologies while maintaining the highest standards of diagnostic quality.

References

1. Boas FE, Fleischmann D. CT artifacts: causes and reduction techniques. Imaging Med. 2012;4(2):229-240. doi:10.2217/iim.12.13
2. Hsieh J. Computed tomography: principles, design, artifacts, and recent advances. 3rd ed. Bellingham, WA: SPIE Press; 2015.
3. Kaza RK, Platt JF, Goodsitt MM, et al. Emerging techniques for dose optimization in CT. Radiographics. 2014;34(1):4-17. doi:10.1148/rg.341135034
4. Lell MM, Kachelriess M. Recent and upcoming technology in computed tomography. Invest Radiol. 2020;55(1):8-19. doi:10.1097/RLI.0000000000000594
5. Mehranian A, Wiegert J, Fiebich M, et al. Quantitative CT metal reduction for patients with hip prostheses. Med Phys. 2021;48(5):2434-2446. doi:10.1002/mp.14789
6. Mohan M, Bindal S, Thawait SK, et al. Metal artifact reduction in CT: where are we after four decades? Radiographics. 2021;41(4):1049-1067. doi:10.1148/rg.2021200192
7. Park C, Yu H, Gupta R, et al. Iterative metal artifact reduction in CT using normalized metal artifact reduction algorithm. Med Phys. 2020;47(7):3216-3228. doi:10.1002/mp.14158
8. Pessis E, Campagna R, Sverzut JM, et al. Virtual monochromatic spectral imaging with fast kilovoltage switching: reduction of metal artifacts at CT. Radiographics. 2013;33(2):573-583. doi:10.1148/rg.332125124
9. Rassouli N, Etesami M, Dhanantwari A, et al. Detector-based spectral CT with a novel dual-layer architecture: principles and potential applications. Eur Radiol. 2017;27(2):498-506. doi:10.1007/s00330-016-4417-1
10. Solomon J, Mileto A, Nelson RC, et al. Quantitative features of liver lesions, lung nodules, and renal stones at multi-detector CT examinations: dependency on radiation dose and reconstruction algorithm. Radiology. 2016;279(1):185-194. doi:10.1148/radiol.2015150807
11. Taguchi K, Iwanczyk JS. Vision 20/20: single photon counting x-ray detectors in medical imaging. Med Phys. 2013;40(10):100901. doi:10.1118/1.4823771
12. Yu L, Li H, Fletcher JG, et al. Automatic CT dose reduction using iterative reconstruction for evaluating suspected renal stones: a prospective evaluation of image quality. AJR Am J Roentgenol. 2016;207(5):935-941. doi:10.2214/AJR.15.15857
13. Barrett JF, Keat N. Artifacts in CT: recognition and avoidance. Radiographics. 2004;24(6):1679-1691. doi:10.1148/rg.246045065
14. Goldman LW. Principles of CT and CT technology. J Nucl Med Technol. 2007;35(3):115-128. doi:10.2967/jnmt.107.044420
15. Kalender WA. Computed tomography: fundamentals, system technology, image quality, applications. 3rd ed. Erlangen: Publicis Publishing; 2011.
16. McCollough CH, Primak AN, Saba O, et al. Dose performance of a 64-channel dual-source CT scanner. Radiology. 2007;243(3):775-784. doi:10.1148/radiol.2433060270
17. Schmidt TG, Fuchs T, Kachelriess M. A review of recent advances in computed tomography. Phys Med. 2020;76:102-115. doi:10.1016/j.ejmp.2020.10.014
18. Söderberg M, Gunnarsson M. Automatic exposure control in computed tomography: an evaluation of systems from different manufacturers. Acta Radiol. 2010;51(6):625-634. doi:10.3109/02841851003757546
19. Yu L, Liu X, Leng S, et al. Radiation dose reduction in computed tomography: techniques and future perspective. Imaging Med. 2009;1(1):65-84. doi:10.2217/iim.09.8
20. Samei E, Li Z, Toth TL, et al. Image quality evaluation in CT: from physical measurements to model observers. Phys Med. 2017;44:72-79. doi:10.1016/j.ejmp.2017.06.053
21. Deak P, Smal Y, Kalender WA. Multisection CT protocols: sex- and age-specific conversion factors used to determine effective dose from dose-length product. Radiology. 2010;257(3):743-749. doi:10.1148/radiol.10100297
22. Mahesh M. The AAPM/RSNA physics tutorial for residents: CT radiation dose and image quality. Radiographics. 2002;22(4):949-961. doi:10.1148/radiographics.22.4.g02jl30949
23. McCollough CH, Bushberg JT, Fletcher JG, et al. Answers to common questions about the use and safety of CT scans. Mayo Clin Proc. 2015;90(10):1380-1392. doi:10.1016/j.mayocp.2015.07.011
24. Slovis TL. The ALARA concept in pediatric CT: myth or reality? Radiology. 2002;223(1):5-6. doi:10.1148/radiol.2231011651
25. Vrtiska TJ, Fletcher JG, McCollough CH. State-of-the-art imaging with 64-channel multidetector CT angiography. Perspect Vasc Surg Endovasc Ther. 2005;17(2):133-142. doi:10.1177/153100350501700205