9.3 Patient-Related Artifacts — Beam Hardening, Motion, Metal & Edge Gradient

Why This Topic Matters

Artifact Recognition and Reduction (2.C) is one of the most heavily tested lettered items inside the Image Evaluation and Archiving subcategory, and artifact-identification scenarios are a hallmark ARRT CT question format: a description of an image appearance is given, and the candidate must name the artifact, its cause, and its fix. This section covers the six sub-items that originate from the patient — the anatomy or positioning being scanned — as distinct from the equipment-based artifacts covered in the next section. Recognizing which category an artifact belongs to matters clinically too: patient-related artifacts are addressed with technique or positioning changes, while equipment artifacts require service intervention.

Beam Hardening (or Cupping)

A clinical CT x-ray beam is polychromatic — it contains a spectrum of photon energies, not a single energy. As this beam passes through tissue, lower-energy photons are absorbed preferentially (the photoelectric effect is strongly energy-dependent), so the beam's average energy increases, or "hardens," the farther it travels through the patient. This produces two classic appearances:

• Dark streaks or bands between two dense structures — the textbook example is a dark band between the petrous bones on a posterior fossa head CT (sometimes called an interpetrous or Hounsfield artifact).
• Cupping artifact in a large, relatively homogeneous structure (a full-size body phantom, or a large patient's abdomen), where the periphery appears artifactually brighter/higher in HU than the center, because central rays traverse more attenuating tissue and harden more.

Correction: software beam-hardening correction algorithms applied at reconstruction, physical prefiltration of the beam (flattening the energy spectrum before it reaches the patient), higher kVp (reduces the relative severity), and dual-energy CT, which can generate monochromatic images that avoid the beam-hardening effect entirely.

Partial Volume Averaging

Partial volume averaging occurs when a single voxel contains a mixture of two or more tissue types with different attenuation coefficients. The reconstructed CT number for that voxel is a weighted average of the tissues inside it — not an accurate representation of either one. It is most problematic when:

• Slices are thick relative to the structure being imaged.
• The structure of interest is small.
• The structure runs obliquely to the scan plane, increasing the length of overlap within a voxel.

A classic clinical trap: a small, truly low-density renal cyst adjacent to the collecting system can appear falsely denser on a thick-slice scan because the voxel averages in higher-density adjacent tissue — potentially masking a stone or a lesion that thinner slices would have revealed clearly. Correction: thinner acquisition and reconstruction slices, isotropic voxel reconstruction, and orienting the scan plane perpendicular to the structure of interest when clinically feasible.

Motion Artifact

Motion artifact results from any patient movement during data acquisition — voluntary (shifting position, coughing, talking) or involuntary (respiratory motion, cardiac pulsation, peristalsis, vascular pulsation). It appears as blurring, streaking, or a double-edge (ghosting) pattern, most visible at high-contrast interfaces such as the diaphragm or aorta. Correction: shorter scan time (faster gantry rotation, wider detector coverage per rotation), clear breath-hold instructions and patient communication, immobilization devices, and — specifically for cardiac motion — ECG gating (prospective or retrospective), which synchronizes acquisition or reconstruction to a defined phase of the cardiac cycle. Adjusting kVp or mAs does not correct motion; those parameters govern noise and dose, not temporal blur.

Metallic Artifact

High-density, high-atomic-number objects — dental amalgam, hip or joint prostheses, surgical clips, pacemaker leads, spinal hardware — cause severe photon starvation (nearly all photons absorbed or scattered before reaching the detector) combined with intense beam hardening. The result is bright and dark streaks radiating outward from the metal, frequently obscuring adjacent anatomy. Correction: dedicated metal artifact reduction (MAR) algorithms (iterative or projection-interpolation based), dual-energy virtual monochromatic imaging at higher keV levels, adjusted technique (higher kVp/mAs when clinically appropriate), and thinner slices. MAR reduces but does not fully eliminate metallic artifact — the exam expects "reduces," not "eliminates," as the correct framing.

Edge Gradient Artifact

Edge gradient artifact arises from undersampling at sharp, high-contrast interfaces — for example, bone-air or bone-soft tissue boundaries at the skull base or petrous ridge — where the system's spatial and contrast resolution cannot fully resolve the abrupt attenuation change. It appears as fine streak lines emanating from the high-contrast edge. Correction: thinner slices, higher-resolution reconstruction algorithms, and adequate sampling (an appropriately small pitch where clinically warranted).

Patient Positioning (Out-of-Field) / Truncation Artifact

When part of the patient's body extends beyond the scan field of view (SFOV) — common with large or bariatric patients, or when a patient's arms remain at their sides during an abdominal/pelvic scan — the reconstruction algorithm's assumption of complete data within the field is violated. Missing peripheral data produces a bright rim or streak artifact at the image edge (truncation artifact), and can also distort CT numbers even in the clinically important central anatomy. Correction: proper patient positioning and centering within the SFOV before scanning (for example, raising the arms above the head for abdominal/pelvic CT whenever possible), selecting an appropriately large SFOV, or applying extended-FOV/truncation-correction reconstruction algorithms when repositioning is not possible.

Exam Scenario

A bariatric patient is scanned for an abdominal/pelvic CT but cannot raise both arms above the head due to shoulder mobility limitations, so the arms remain at the patient's sides within the gantry. The resulting image shows streak artifact concentrated at the lateral image periphery, while the central abdominal anatomy is comparatively less affected. The technologist should recognize this as an out-of-field (truncation) artifact related to the patient's arms extending beyond the SFOV — not a scanner malfunction — and should consider selecting a larger SFOV or an extended-FOV reconstruction algorithm rather than assuming the equipment needs service.