Equipment & Patient Artifacts: Ring, Motion, Metal & Windmill Effects

Identify the source before choosing a correction

An artifact is a reconstructed feature that does not accurately represent the anatomy or attenuation distribution. Equipment calibration, patient motion, metal, incomplete data and helical geometry can all contribute. A name alone is not a troubleshooting plan: first relate the appearance to the scanner, patient and acquisition. Additional exposure is justified only when the responsible team determines that information cannot be obtained from the existing data.

Rings and calibration errors

In a rotate-rotate scanner, a detector channel samples rays at a consistent distance from the rotation center. A channel gain or offset error can therefore create a ring or arc centered on isocenter. The affected radius and slices depend on the channel, detector row, data processing and acquisition. Not every failed row maps to exactly one clinical slice because helical reconstruction can combine several rows and views.

A ring in a properly acquired uniform phantom supports a system or calibration problem rather than patient pathology. Perform manufacturer-authorized calibration and repeat the approved QC test. Persistent errors require escalation to the physicist or service team. Do not make daily air and water calibration a universal requirement for every manufacturer, or confuse calibration with the daily independent water test. Save the artifact and settings so that service can reproduce it.

Motion creates inconsistent projection information

The reconstruction assumes a consistent object during the data used for an image. Breathing, swallowing, cardiac pulsation or gross movement can change the anatomy between views, causing blurring, duplicated edges, shading or streaks. Movement between adjacent acquisitions can create discontinuities in reformations.

Coach and practice a feasible breath hold, explain table motion and contrast warmth, and use safe supports. Preserve trauma immobilization and avoid restraint that impairs breathing or causes injury. Faster acquisition or appropriate gating can help, but the fastest rotation and highest pitch are not automatically the best settings for every diagnostic task. Temporal resolution, output limits and sampling must also remain suitable.

Cardiac motion at the ascending aorta can mimic a flap. ECG-synchronized imaging may reduce that ambiguity under the approved protocol. Prospective triggering can target systolic or diastolic windows depending on patient and scanner; it is not exclusively a late-diastolic method. Ordered medications or sedation require appropriate screening, personnel and monitoring and do not guarantee complete motion suppression.

Metal combines several mechanisms

Metal can produce beam hardening, photon starvation, scatter, partial-volume and nonlinear effects. Some paths may have extremely few transmitted photons, but do not assert that every metal path has exactly zero counts. A very large attenuation value is a reconstructed image quantity, not itself the detector's input voltage or proof of ADC saturation.

Remove avoidable external metal and position implants outside the critical plane when clinically feasible. Use an appropriate acquisition and reconstruction for the task. Higher voltage or output may improve penetration but changes dose; simply selecting the maximum available values is not a general protocol. Thin data, selected high-energy virtual monoenergetic images and metal-artifact-reduction (MAR) algorithms can help.

MAR methods may identify metal-corrupted projections and estimate replacement information. They can reduce streaking but may introduce new artifacts or alter structures near the implant. Compare conventional and MAR images when needed. High-keV virtual monoenergetic imaging also has limitations: it is a synthesized material-based image, not a guarantee of a physically monochromatic acquisition or complete beam-hardening removal. It may reduce iodine conspicuity as energy rises.

Helical and cone-beam effects

Multidetector arrays sample rays that diverge along the longitudinal direction. Helical sampling near sharp longitudinal changes can create rotating bright and dark streaks often called windmill artifacts. Pitch, detector geometry and reconstruction influence the effect. The presence of many rows does not mean every scanner uses the same cone-beam reconstruction or experiences the same severity.

Use the manufacturer's validated reconstruction and protocol. Changing pitch or using axial acquisition can help selected tasks, but a pitch below 1.0 is not a universal solution. Flying focal spots improve sampling in supported designs without automatically halving all cone-angle errors. Wide axial acquisitions can still have cone-beam and other artifacts even though table translation is absent.

Truncation and the fields of view

The scan field of view (SFOV) describes the acquired and calibrated measurement region. The display or reconstruction field of view (DFOV) describes the reconstructed area. If attenuating anatomy extends beyond the measured region, missing projection information can cause shading, streaks and inaccurate CT numbers. Selecting a small DFOV from otherwise complete acquired data is different from physical truncation.

Center the patient and place arms above the torso when safe and appropriate. If that position is unsafe, use an approved alternative and document the limitation. Selecting the appropriate body SFOV may help, but a larger setting cannot make an undersized physical detector measure every missing ray. Extended FOV methods estimate information beyond the standard region; image quality and attenuation accuracy there may differ from those within the measured region, especially for quantitative or radiation-planning use.

Apply the distinction in practice

Suppose a hip prosthesis obscures the adjacent pelvic soft tissue. Increasing dose may improve photon statistics but will not necessarily remove nonlinear metal effects. Review the available MAR and spectral reconstructions and ask whether they restore the clinically needed information. If a new dark band appears only on MAR, compare the conventional image before calling it disease.

Alternatively, a faint ring appears in multiple patients and in the water phantom at the same radius. Patient coaching is unlikely to solve it; the evidence supports a system investigation. These scenarios illustrate why an artifact's cause should determine the response and why the final record should include the limitation and corrective action.

References: AAPM CT protocols and terminology, ACR CT QC program.