Image Production · Digital imaging and processing3 min read5-question check
Image Artifacts, Saturation and Virtual Grids
An artifact in the same spot on every patient is the detector; a wavy pattern is the grid; a black patch with no detail is saturation.
Patient and equipment
An artifact is anything on the image that is not part of the patient's anatomy, and it can hide or mimic pathology. Patient artifacts include jewelry, hair clips, gown snaps, braids, clothing, and motion. Remove what you can, and note what cannot be removed, such as an implant or monitoring line. Equipment artifacts come from the detector or the tube. Dead pixels or dead lines appear in the same place on every image from that detector, whatever the patient. Ghosting, or lag, is a faint leftover image from a previous high exposure that the detector has not fully cleared. Detector cracks, debris in the collimator, and calibration errors also repeat from patient to patient. Those need recalibration or service, not a repeat.
Grid artifacts
With a stationary grid, a moiré pattern, a wavy interference pattern, can form when the grid lines run parallel to the detector's readout lines and the grid frequency is close to the detector's sampling frequency. Prevent it with a high-frequency grid, a moving Bucky grid, or a grid turned so its lines run perpendicular to the readout direction. Grid lines visible on the image usually mean a damaged or stationary low-frequency grid. Light edges with normal density in the center point to grid cutoff from an upside-down focused grid or an SID outside the grid's focal range. An off-level or off-center grid causes a more uniform loss of density instead.
Exposure errors
Gross exposure errors are artifacts too. Underexposure leaves the image grainy with quantum noise, even after rescaling makes the brightness look right. Severe overexposure can saturate detector elements. Saturated areas turn black with no data at all, often at the skin line or thin edges, and contrast drops. No processing can bring back saturated data. If an area of interest is saturated or too noisy to read, the exposure must be repeated with a corrected technique.
Virtual grids
A virtual grid is software, not a physical grid. It estimates how much scatter is in the image from the kVp and part thickness and subtracts that effect, raising contrast. It is used mostly for bedside and mobile work where a grid is awkward to align. Because there is no lead in the beam, there is no grid cutoff, no grid lines, no moiré, and no Bucky factor, so dose can stay lower than with a real grid. Scatter still reaches the detector, though, so a virtual grid improves contrast but does not remove scatter physically.
Clinical Example
Three different patients imaged in room 4 all show the same faint vertical white line near the left edge. The radiographer realizes it is a detector problem, not a patient one. She takes the detector out of service, runs the calibration, and reports it when the line persists.
Common Mistakes
- Repeating the patient for an artifact that is on every image from the same detector.
- Assuming a saturated area can be brightened in processing.
- Thinking a virtual grid physically blocks scatter.
- Blaming moiré on patient motion.
Why It Matters
Image evaluation, artifacts, saturation, and virtual grids are all named in the 2027 Image Production outline, and virtual grids are new. Artifact questions usually describe what you see and ask for the cause or the fix.
Key Relationships
- Same spot on every patient → detector defect
- Faint prior image → ghosting or lag
- Wavy pattern with stationary grid → moiré
- Black area, no detail → saturation; repeat needed
- Virtual grid → software scatter correction, no cutoff, no Bucky factor