What you see

The deviation index (DI) or the vendor’s exposure index is outside your department’s range, although the image may look normal; or the image looks noisy.

Why it happens

The exposure index (EI) measures the signal at the detector for the image; the target EI is set per body part and view, and the DI shows how far this exposure is from the target (0 on target, positive above, negative below). Processing makes over- and underexposed images look similar in brightness, so an overexposed image can look good while giving the patient more dose than needed (dose creep). An underexposed image looks noisy.

Try first

1. Check you chose the right body part and view: the wrong selection uses the wrong target and gives a wrong DI.
2. Check whether something upset the reading: tight or uneven collimation, a prosthesis, a shield in the field, unusual positioning, or (on CR) a long delay before reading the plate.
3. Low DI: zoom and pan to judge noise on a good display; a noisy image may need a repeat after review.
4. High DI: do not repeat because the DI is high. Repeat an overexposed image only if relevant anatomy is saturated (clipped), see Saturation.
5. Correct the technique for the next image, and log extreme DI values for review.

Stop and escalate when

The DI is far out of range without an obvious cause, or out of range on many images: tell the lead radiographer and medical physicist (exposure, AEC or calibration may need checking).

Where sources differ

Action limits. AAPM TG-116 (2009) gave a target DI range of -0.5 to +0.5, with a repeat below -3 and management follow-up above +3. AAPM TG-232 (2018, abstract) found these too strict for clinical practice and recommends site-specific action limits at 1 and 2 standard deviations of each site’s own DI data; the ACR practice parameter (2022) says each practice sets its own limits and an extreme DI alone should not prompt an immediate repeat. Follow your department’s limits.

Sources: AAPM TG-116 2009; Seibert 2011; AAPM TG-232 2018; ACR DR PP 2022; AAPM Report 93; Mothiram 2014; Kowalczyk 2009; Schaefer-Prokop 2008

What you see

The image looks grainy or speckled, especially in thick or dense regions; fine low-contrast detail is hard to see.

Why it happens

Noise comes mainly from the random arrival of X-ray photons: too few photons reach the detector. Causes include too low an exposure for the patient’s size, the wrong technique or AEC set-up, or an image displayed on a monitor that hides or exaggerates noise.

Try first

1. Check the DI (see Exposure index) and the technique against the protocol for the patient’s size.
2. Judge noise on a diagnostic-quality display, zoomed in; acquisition and QC monitors are often not calibrated and can mislead.
3. If AEC was used, check the right chambers were selected and covered by the anatomy (see AEC problems).
4. If the noise hides what the request needs, a repeat with corrected technique may be justified after review.

Stop and escalate when

Noise on images of every patient with an unchanged technique: report it for a QC check of the detector and AEC.

Sources: IAEA DRP 2014; AAPM TG-116 2009; Seibert 2011; ACR DR PP 2022

What you see

Areas of the image show no detail at all (burned out or clipped), often after a high exposure.

Why it happens

When the detector or its electronics receive more signal than they can record, every pixel in that area is forced to the maximum value, so the information there is lost and cannot be recovered by processing.

Try first

1. Check the technique and the DI: saturation follows a large overexposure.
2. Check the right body part, view and processing were chosen; wrong processing can also make areas look washed out (see Processing).
3. If relevant anatomy is saturated, a repeat with corrected technique is justified; it is the main reason to repeat an overexposed digital image.

Stop and escalate when

Saturation at ordinary exposures: report it, as the detector or reader may need checking.

Sources: AAPM TG-116 2009; AAPM Report 93; ACR DR PP 2022; Mothiram 2014

What you see

Fine parallel lines across the image, or a coarse wavy interference pattern (Moiré), sometimes visible only on thumbnails or at certain zoom levels.

Why it happens

A stationary grid’s lead strips can show as lines, and their regular pattern can interfere with the detector’s pixel grid or the CR reader’s scan lines, producing Moiré. Moving (reciprocating) grids blur the strips out. A damaged grid also causes artefacts.

Try first

1. Check the image at full size: Moiré that appears only on a reduced thumbnail is a display effect.
2. If a stationary grid was used, check it is the type your system specifies for that detector; on CR, grid lines should run across the reader’s scan direction.
3. Check the Bucky grid moved during the exposure (see Bucky and grid).
4. Inspect the grid for dents or damage.

Stop and escalate when

Lines appear with the correct grid, or the grid is damaged: take the grid or Bucky out of use and report it.

Sources: IAEA DRP 2014; AAPM Report 93; AAPM TG-150 2024; Hegde 2018

What you see

Part of the image has less signal than it should: more noise or lower density towards one or both sides, or across the whole image, especially on bedside chest images taken with a grid.

Why it happens

A focused grid’s strips are angled to match the beam at one distance. At the wrong distance (off-focus), the sides lose signal while the centre is unchanged; off-centre to the grid (lateral decentring), signal falls evenly across the whole image. A slightly tilted grid acts like lateral decentring; a larger tilt, or being off-centre and at the wrong distance together, makes the loss uneven, with one side noisier or lighter than the other. A grid used upside down also causes cut-off. Mobile grids are often hard to align, and in bedside chest work a tilt of more than 3 degrees can lead to interpretation errors (Fischer).

Try first

1. Check the source-to-image distance matches the grid’s focal range marked on the grid.
2. Centre the beam to the grid, and keep the beam at right angles to the grid; for bedside work, check the grid is not tilted under the patient.
3. Check the grid is the right way up.
4. Consider whether a grid is needed for that patient and view.

Stop and escalate when

Cut-off on fixed equipment at the correct distance and centring: report it, as a detent, the grid mounting or the grid itself may be faulty.

Can mimic pathology

Uneven cut-off can change the apparent darkness of one lung and imitate disease; compare areas outside the chest.

Sources: IAEA DRP 2014; Gauntt 2010; Fischer 1992; EUR 16260 1996; AAPM TG-150 2024; AAPM Report 93

What you see

Signal falls off along the length of a large field, towards the anode end, for example along a femur, spine or long-length image.

Why it happens

X-rays produced deep in the anode are absorbed more on the anode side, so the beam is less intense there. The effect is often barely noticeable in radiographs; in digital systems post-processing is assumed to compensate for much of it, although it may also alter the image’s appearance.

Try first

1. For long fields, use the SID and field size your protocol sets.
2. Follow your department’s positioning protocol for tube orientation; the evidence for a particular orientation in digital radiography is inconclusive (below).

Good to know

A systematic review (Kusk and colleagues, 2021) found that the evidence neither proves nor disproves an effect on image quality in digital radiography; one phantom study of the AP pelvis found lower gonad dose with the feet towards the anode, with visual image quality unchanged but signal-to-noise changed.

Sources: IAEA DRP 2014; Kusk 2021; Mraity 2020

What you see

Edges are blurred or doubled; on the chest, the heart borders or diaphragm are blurred; on long exposures, fine bone detail is lost.

Why it happens

The patient or the part moved during the exposure: breathing, pain, tremor, a child or a confused patient. Longer exposure times, for example with a very small focal spot or low kV, raise the risk.

Try first

1. Explain the breathing instruction and practise it; give the instruction and expose at the right moment.
2. Support the part with pads, sandbags or immobilisation devices so the patient does not have to work to keep still.
3. Use a shorter exposure time where your protocol allows (the right focal spot and kV for the region).
4. For a child or a patient who cannot keep still, see Children and Cannot stand or hold position.

Stop and escalate when

Motion keeps spoiling images: ask the radiologist which view matters most before repeating.

Sources: IAEA DRP 2014; ACR DR PP 2022; Schaefer-Prokop 2008; Atkinson 2020; EUR 16260 1996; ASMIRT PPS 2018

What you see

Another image, outline or object faintly visible on a CR image: a previous patient’s anatomy, a lead marker, or the outline of a high-exposure area.

Why it happens

The reader removes only part of the stored signal; the plate is then erased with bright light. Erasure can be incomplete after a very high exposure, or the plate may not have been erased before reuse, or it was exposed twice. Background radiation also builds up on plates left unused.

Try first

1. Erase plates before use if you are unsure of their status, and use one exposure per plate.
2. After a very heavy exposure, erase the plate again (possibly more than once) before reuse.
3. Erase infrequently used plates on the schedule your QC programme sets.

Stop and escalate when

Ghosts persist after erasure, or appear on many plates: take the plates or reader out of use and report it.

Can mimic pathology

A ghost from a previous image has imitated a fetus in the abdomen on a chest film; residual images can imitate disease.

Sources: AAPM Report 93; IAEA DRP 2014; Hegde 2018; IAEA HHS 47

What you see

A faint residual image of a previous exposure, an object or a collimated edge on a later DR image.

Why it happens

Lag is signal carried over into the next image; ghosting is a change in detector gain left by a previous image. Both are worst after very large exposures, heavy exposure of lead markers, or repeated small collimated fields.

Try first

1. Note the time, the previous exposure and what showed.
2. Allow time before the next exposure if your manufacturer advises it, and avoid unnecessary very high exposures to the detector.
3. Run the detector’s calibration if it is a user task on your system and the manufacturer recommends it.

Stop and escalate when

The ghost persists or appears on clinical images: report it to service and the medical physicist.

Sources: IAEA DRP 2014; AAPM TG-150 2024; ACR DR PP 2022; EC RP 162

What you see

White specks or spots, fine lines or scratches in the same place on images from one plate; a straight line across images from one reader; image signal lower than expected after a delay.

Why it happens

Dust and dirt on the plate, scratches or damage to the phosphor, dust in the reader’s light-collection optics, and fading of the stored image when the plate is left before reading (a typical plate loses about a quarter of its signal between 10 minutes and 8 hours after exposure).

Try first

1. Identify the plate (the artefact follows the plate) or the reader (the artefact appears on every plate).
2. Clean the plate only with the manufacturer’s approved method; erase it and re-check.
3. Read plates promptly after exposure.
4. Withdraw a plate that cannot be cleaned or is damaged.

Stop and escalate when

Lines appear on every plate through one reader: take the reader out of use and call service.

Sources: AAPM Report 93; IAEA HHS 47; IAEA DRP 2014; Kowalczyk 2009

What you see

A fixed line, cluster of dead pixels, rectangular blocks or non-uniform areas in the same place on every image from one detector, or an outline of an object that was not there.

Why it happens

Detectors are corrected for faulty elements and uneven gain by a calibration map. The map can become less effective over time, calibration can be done wrongly (an object left in the beam during calibration is burned into every later image), and detectors can be damaged, for example by a drop.

Try first

1. Check the same defect on a flat-field or QC image.
2. Run the detector’s calibration if it is a user task on your system, as the manufacturer specifies, with nothing in the beam.
3. Note the detector’s identity, especially with several wireless detectors.

Stop and escalate when

The defect persists after calibration, or the detector was dropped: take it out of use, report it, and tell the radiologists so they can judge its effect on clinical images.

Sources: IAEA HHS 47; ACR DR PP 2022; AAPM TG-150 2024; ACPSEM 2024; Willis 2011; Walz-Flannigan 2018

What you see

A step, gap, overlap or false line where sections of a long-length (whole spine or leg) image are joined.

Why it happens

Long-length images are made from several exposures joined by software, and an error in the merging step misaligns the sections. In one reported case the merging error made a normal vertebra look fractured.

Try first

1. Use the system’s long-length method as it is set up; the ACR expects the equipment to join images without exposing sections twice.
2. Keep the patient still between exposures, with the positioning aids your protocol specifies (a general step).
3. Check the joins on the stitched image before the patient leaves.
4. Label stitched images and mark the overlap regions, as Benes and colleagues recommend, so readers know where joins are.

Stop and escalate when

A join looks like a fracture or a broken device: tell the radiologist; focused images of that level may be needed.

Can mimic pathology

Stitching errors have imitated a vertebral fracture and a broken shunt catheter.

Sources: Benes 2023; Hodnett 2025; ACR DR PP 2022

What you see

A faint image of an object or structure that lies behind the detector, such as part of the table, holder or cassette.

Why it happens

X-rays passing through the detector scatter back from material behind it. Cassettes and detectors have backing to limit this, and some older CR cassettes lacked adequate backscatter control.

Try first

1. Check what was behind the detector during the exposure, and remove it or use the correct holder.
2. Collimate to the area needed, which reduces scatter.

Stop and escalate when

The pattern appears with the correct set-up: report the cassette or detector for checking.

Sources: IAEA DRP 2014; AAPM Report 93; Walz-Flannigan 2018

What you see

The image looks washed out, too light or too dark, with poor contrast, or has halos around metal or edges, although the exposure was right.

Why it happens

Digital images are processed for the body part chosen. The wrong examination or processing menu, a failed recognition of the collimated area (histogram analysis), or strong edge enhancement can mis-scale the image; edge enhancement can create halos.

Try first

1. Check the examination and view selected match the image.
2. Reprocess with the right menu, or correct the collimation or region of interest, rather than repeating the exposure.
3. Do not change processing parameters yourself beyond the options your department allows; changes are made with the applications specialist and radiologists.

Stop and escalate when

Many images look wrong after a software change: report it to the lead radiographer and physicist.

Sources: ACR DR PP 2022; AAPM Report 93; Schaefer-Prokop 2008

What you see

Part of the anatomy requested is missing at the edge of the image, or the collimated edges are not visible because the image has been cropped or electronically masked.

Why it happens

Collimating to the area of interest lowers dose and improves the image, but collimating too tightly cuts off anatomy and leads to a repeat. Cropping after exposure does not reduce the area irradiated, and electronic masking can hide anatomy or poor collimation. Anatomy cut-off is one of the commonest reasons for rejected images.

Try first

1. Check the area requested and collimate to it with the light field, allowing for the part’s full extent.
2. Keep the physical collimation visible as a thin border; do not use electronic masking to hide anatomy or wide collimation.
3. Check the light field matches the X-ray field (see Collimator light).

Stop and escalate when

Anatomy is cut off and the image does not answer the question: ask the radiologist whether a repeat is needed.

Sources: IAEA DRP 2014; ACR DR PP 2022; IAEA SSG-46; Atkinson 2020; Stephenson-Smith 2021; Hayre 2019

What you see

The side marker is missing, on the wrong side, or added only electronically; or it hides anatomy.

Why it happens

The side shown on an image must be unambiguous. A marker placed in the beam before exposure records the side at the time; markers added afterwards can be placed wrongly, and images can be flipped.

Try first

1. Place an anatomical side marker in the beam, inside the collimated field, before every exposure, clear of the anatomy.
2. Check the marker on the image matches the side requested and the patient.
3. If the marker is missing but right and left are absolutely clear from the anatomy, a repeat is not normally needed; if there is any doubt, repeat.

Stop and escalate when

An image with the wrong side has been sent: tell the radiologist straight away and correct it through your authorised process; report it as an incident.

Where sources differ

Lead versus digital markers. The SCoR (UK), the CAMRT (Canada) and the RCR audit standard expect a marker in the beam at exposure, with post-processing markers only a safety net; ASMIRT’s standards (Australia) say to verify that markers are present on each image and accurate, without specifying the type; the ACR practice parameter requires the side to be shown on every image without specifying how; a UK study found many radiographers treat digital and anatomical markers as interchangeable.

Sources: SCoR markers 2014; CAMRT BPG; RCR XR-503; ACR DR PP 2022; Chung 2020; Hayre 2019; ASMIRT PPS 2018

What you see

Buttons, fasteners, wires, folds of clothing, bed sheets, skin folds or positioning aids visible on the image; curved lines over the chest.

Why it happens

Anything in the beam is imaged. Clothing artefacts are a common reason for rejected images, and some positioning aids show on digital detectors.

Try first

1. Remove clothing and objects from the area before the exposure, as your department’s gown policy describes.
2. Smooth sheets and skin folds where you can, and check the image before the patient leaves.
3. Tell the radiologist about an object that could not be removed.

Can mimic pathology

Skin folds, sheet folds and clothing can imitate a pneumothorax; lung markings continuing beyond the line favour a fold.

Sources: Stephenson-Smith 2021; Jalamneh 2023; ACR DR PP 2022

What you see

The image is rotated or tilted, for example on a PA chest the spinous processes are not midway between the medial ends of the clavicles, the scapulae overlie the lungs, or vertebral end plates are not seen as single lines.

Why it happens

Positioning is the commonest reason for rejected digital images. Quality criteria for standard projections describe what a well-positioned image shows.

Try first

1. Check the patient’s position and the centring against your projection protocol before exposing, using the quality criteria for that view.
2. Use positioning aids so the patient can hold the position without effort.
3. Judge the image against the clinical question: an image that answers the question is not rejected only because a criterion is missed.

Stop and escalate when

The image cannot answer the question: ask the radiologist before repeating.

Good to know

The European quality criteria for a PA chest include full inspiration and symmetry shown by the spinous processes lying centrally between the medial ends of the clavicles. In one Australian emergency department, horizontal-beam views were among the projections with both many rejects and high reject rates: 33.5% for the hip and 30.5% for the knee (Stephenson-Smith and colleagues).

Sources: EUR 16260 1996; Atkinson 2020; Stephenson-Smith 2021; Hofmann 2015

What you see

The DR detector will not show ready, is not recognised by the console, loses its wireless link, runs out of battery, or an exposure is made but no image appears.

Why it happens

The console must have the right detector selected and initialised before the exposure; a Bucky not pushed fully home may not initialise its detector. Wireless detectors depend on their link and battery; images can stay in the detector’s memory when transfer fails. Reported faults across manufacturers include detectors falsely showing ready, battery overheating, and an image left in a detector’s memory after a weak wireless link, which was lost when the detector was reset by removing its batteries (FDA recall Z-0704-2018).

Try first

1. Before exposing, check the console shows the right detector as selected and ready, and the Bucky or holder is fully home.
2. Check the battery charge and swap to a charged battery between patients, as the manufacturer describes (a general step).
3. For a wireless detector, keep it within range of its access point and follow the manufacturer’s procedure to re-pair it (a general step).
4. If an exposure produced no image, do not remove the battery or restart the detector until you have followed the manufacturer’s image-recovery procedure: the image may still be in the detector’s memory.
5. Record the exposure even if no image appeared; the patient received the dose.

Stop and escalate when

The detector fails repeatedly, a battery is hot, swollen or damaged, or images are lost: take the detector or battery out of use and report it.

Sources: AAPM TG-305; ACR DR PP 2022; AAPM TG-150 2024; FDA recalls

What you see

The console shows a tube or generator error, a heat warning or a cooling wait, an exposure stops early, or output seems to have changed.

Why it happens

The generator calculates tube loading and cooling times from the tube’s rating data; repeated high-power exposures without the required cooling can damage the anode and bearings. Tubes age, and a failing vacuum can damage the filament. Evidence of arcing is a reason to replace a tube.

Try first

1. Note the message, the time and what was being done, and record it in the fault log.
2. Let the tube cool if the console asks; do not try to bypass a cooling wait.
3. Run the daily warm-up the manufacturer specifies before the first patient.
4. Do not keep exposing patients through repeated errors.

Stop and escalate when

Errors recur, an exposure ends early or the image quality changes: stop using the unit, report it, and call service. It is checked before it is used for patients again.

Sources: IAEA DRP 2014; AAPM TG-150 2024; Manitoba DR QC 2019; AAPM Report 74; ARPANSA RPS C-5

What you see

The light field is dim or does not come on, has shadows from dust, or the X-ray field on the image does not match the light field (anatomy cut off on one side, or exposure outside the area set).

Why it happens

The light field shows where the X-ray beam will go. The lamp can fail or dim, and the mirror or alignment can shift, especially on mobile units, where physical knocks can cause misalignment. If they do not match, anatomy is cut off or tissue outside the area is irradiated.

Try first

1. Check the light comes on and is bright enough; clean the collimator window as the manufacturer allows.
2. If the light has failed, the collimator’s field-size scale can be used to set the field until it is repaired, as your local rules allow.
3. Run your QC programme’s light and X-ray field alignment check if misalignment is suspected, and record the result.

Stop and escalate when

The fields do not match beyond your programme’s limit, or the light is out: report it to service and the physicist. IAEA HHS 47 gives both a 30-day time frame for an inoperative light and repair of inoperative or misaligned items before further patients are imaged, so your local rules decide whether the room stays in use.

Where sources differ

Alignment limits differ: ACPSEM (Australia and New Zealand, 2024) recommends within 1% of the focus-to-image distance; the US standard (21 CFR 1020.31) allows a total of 2% of SID along each axis, and IAEA HHS 47 uses 2 cm at 100 cm on any side; the EU criteria (RP 162) take equipment out of use beyond 3%.

Sources: IAEA HHS 47; ACPSEM 2024; 21 CFR 1020.31; EC RP 162; AAPM TG-150 2024; AAPM Report 74; Brookfield 2015

What you see

The Bucky tray or detector will not lock or initialise, grid lines appear, signal falls off at the edges or across the image, or the wrong grid is in place.

Why it happens

A grid must suit the detector and be used within its focal distance; a tray not pushed fully home may not initialise the detector; a mispositioned detent or a tilted or damaged grid causes cut-off; a moving grid that stops leaves visible lines.

Try first

1. Push the tray or detector fully home and check the console shows it ready.
2. Check the grid fitted is the one for that receptor and SID, and is the right way up.
3. Look for grid lines or uneven signal on the image (see Grid lines and Grid cut-off).

Stop and escalate when

Grid lines from a moving grid, cut-off at the correct set-up, or a damaged grid: take the Bucky or grid out of use and report it.

Sources: AAPM TG-150 2024; AAPM TG-305; IAEA DRP 2014; EC RP 162

What you see

AEC images are noisy or the DI is far off target, the exposure ends on the backup timer with a warning, or exposures differ for similar patients.

Why it happens

The AEC measures the radiation through the selected chambers. It gives the wrong exposure if the anatomy is not centred over the active chambers, the wrong chambers are selected, a dense object such as a hip prosthesis or a shield covers a chamber, or a small child’s anatomy only partly covers it. If the AEC does not end the exposure, the backup timer stops it.

Try first

1. Check the selected chambers match the protocol for the view and are covered by the anatomy.
2. If metal or a shield will lie over a chamber, use a size-based manual technique instead.
3. After a backup-timer stop, the system needs resetting; check the set-up before exposing again, and do not repeat at the same settings without finding the cause.
4. Never let a shield cover an active AEC chamber.

Stop and escalate when

The AEC ends on the backup timer with a correct set-up, or exposures are inconsistent: stop using AEC on that unit until it is checked, and report it.

Sources: ACR DR PP 2022; AAPM TG-150 2024; 21 CFR 1020.31; IAEA HHS 47; ACPSEM 2024; NCRP 13; Chen 2026

What you see

A mobile X-ray unit will not drive, drives erratically or moves on its own, does not stop when the drive handle is released, will not charge, or its arm or column is loose.

Why it happens

Manufacturers have recalled mobile units for brakes that did not apply when the handle was released, uncommanded movement (including from static discharge and from cleaning liquid shorting a button), wheel shafts breaking when driven over high thresholds, loose tube-arm bolts, and battery fire risk while charging.

Try first

1. Releasing the drive handle should stop the unit; if it does not, use the emergency stop and keep people clear (a general step; know where your unit’s emergency stop is).
2. Drive slowly over thresholds and into lifts, and keep feet and other people clear.
3. Clean the unit as the manufacturer instructs, without spraying liquid directly onto it.
4. Charge the unit where your department specifies and report a battery that is hot or damaged.

Stop and escalate when

Any uncommanded movement, brake failure, loose arm or charging fault: take the unit out of use, label it, and report it to service and through your incident system.

Sources: FDA recalls; BfArM 2013; FDA MAUDE; ACPSEM 2024

What you see

The dose-area product (DAP or KAP) is not displayed, is zero, differs between the console and PACS, or is far from usual for the examination.

Why it happens

The DAP meter measures the radiation leaving the collimator and is recorded with the image (in the DICOM header and dose report). A missing or mismatched value can come from a meter fault or a set-up fault in how dose data are passed on.

Try first

1. Check the examination and technique were as usual, and that the reading is not explained by a larger field or repeat.
2. Record the technique factors so dose can be estimated if needed.
3. Report a missing or mismatched value so the meter and data transfer can be checked.

Stop and escalate when

The meter is faulty: report it. IAEA HHS 47 says the unit may be used until the meter is repaired if everything else performs, so long as your local rules allow it.

Sources: IAEA HHS 47; AAPM TG-150 2024; ACPSEM 2024; IAEA DRP 2014

What you see

Images do not reach PACS, are missing after transfer, an image is lost after the system froze, or the worklist does not show the patient or shows wrong entries.

Why it happens

The modality sends images to PACS and, with storage commitment, deletes its copies only after PACS confirms storage. Links between a wireless detector and its console often lack that confirmation, so departments need a recovery strategy. Patient details come from the modality worklist; manual entry is error-prone.

Try first

1. Check the send queue and the destination, and resend failed images; compare image counts.
2. Do not delete images from the console until PACS has confirmed storage.
3. If an image appears lost, check the detector and console before re-exposing (see Detector problems).
4. If the worklist is empty or wrong, check the booking, and enter details by hand only under your downtime procedure.
5. Keep rejected images retrievable for reject analysis; do not delete them directly from PACS.

Stop and escalate when

Images are still missing, or a patient identity error is found: contact your PACS or IT team and report it through your incident process.

Sources: DICOM PS3.4; Clunie 2014; ACR DR PP 2022; AAPM TG-150 2024; AAPM TG-305; Martin 2017; FDA recalls

What you see

A routine QC test fails (uniformity, artefacts, exposure index constancy, light field), the system asks for detector calibration, or the reject rate is unusually high or low.

Why it happens

QC catches drift in the tube, AEC, detector and DAP meter before it affects patients. Detector calibration is done as the manufacturer specifies; on some systems it is a radiographer task, on others a service task. Excess non-uniformity or lines can mean recalibration is needed.

Try first

1. Repeat the test with the set-up checked (the same phantom, distance, detector and settings).
2. Run the detector calibration if it is a user task on your system, as the manufacturer specifies, with nothing left in the beam.
3. Record results and actions in the QC log.
4. Review rejected images with their reasons, not only the overall rate.

Stop and escalate when

The test still fails: tell the medical physicist, who decides with service whether the equipment stays in use.

Where sources differ

Reject-rate targets. AAPM TG-151 (via TG-305) suggests 8% for adults with investigation below 5% or above 10%; ACPSEM (Australia and New Zealand) considers 8 ± 2% appropriate for a typical service and 5 ± 2% for paediatric radiography; IAEA HHS 47 reports 4% to 8% and says to set targets locally. TG-305 counts per exposure, ACPSEM per image. A very low rate can mean rejects are not being recorded.

Sources: ACPSEM 2024; IAEA HHS 47; AAPM TG-150 2024; AAPM TG-305; ACR-AAPM Radiographic TS 2026; Manitoba DR QC 2019; ACR DR PP 2022

What you see

Before exposing, the identity of the patient or the side to be imaged needs confirming: a patient who cannot answer, a common name, a ward patient, or a request that does not match what the patient says.

Why it happens

Wrong-patient and wrong-side exposures are always possible. At least two, preferably three, identifiers checked against the request, open questions, and the side confirmed with the patient prevent most of them. Images must show the patient’s identity and the side imaged.

Try first

1. Ask the patient to state their full name and date of birth (open questions) and what they are having done, including the side; match the answers to the request and identity band. A room or bed number is not an identifier.
2. For a patient who cannot answer, use the identity band, the health record or the staff caring for them, as your policy describes.
3. Select the right patient, examination and projection on the console (for example AP or PA), since this sets the image labels.
4. Place a side marker in the beam (see Wrong or missing marker) and check details and side on the image before release.

Stop and escalate when

Any mismatch: do not expose until it is resolved, and report it through your incident process.

Sources: IAEA SSG-46; CAMRT BPG; ORS C1 2024; ACR DR PP 2022; ASMIRT PPS 2018; Chung 2020

What you see

A patient of childbearing potential is booked for an examination that includes the abdomen or pelvis, and might be pregnant, is not sure, or the answer is unclear; or a pregnant patient is booked.

Why it happens

Pregnancy status is checked before exposures where the primary beam may irradiate the pelvis. Most plain radiographs give the embryo or fetus a very small dose: chest, extremity and head radiographs give it only scattered radiation, except that a chest radiograph late in pregnancy can include part of the fetus in the beam (the ACR-SPR parameter suggests a frontal view only), and the ACR-SPR parameter says most radiographic examinations deliver much less than 20 mGy to a conceptus.

Try first

1. Ask every patient of childbearing potential, privately, using your department’s questions or form, immediately before the examination, and record the answer.
2. If pregnancy cannot be excluded for an examination of the abdomen or pelvis, follow your local procedure: the radiologist or practitioner decides whether to proceed, delay or change the examination.
3. For a pregnant patient whose examination is justified, collimate tightly to the area needed; prone positioning for pelvic examinations reduces fetal dose where possible (Health Canada). The CAMRT page on pregnant patients, which is under review, also advises PA rather than AP positioning where possible.
4. Lead shielding of the abdomen for examinations of other regions is not required by current UK and US guidance; a drape may be offered for reassurance if your written procedure allows it (see Shielding requests).

Stop and escalate when

The patient is or may be pregnant and the abdomen or pelvis is in the beam: the practitioner or radiologist decides before the exposure.

Where sources differ

When to ask differs by region. UK (HPA 2009, RCR 2026): ask when the primary beam may irradiate the pelvis, between the diaphragm and the knees; for low-dose examinations (below about 10 mGy fetal dose, which covers plain radiographs) proceed if the period is not overdue. US (ACR-SPR 2023): chest, extremity, head and neck radiography generally need no verification. Australia (ARPANSA RPS C-5): establish status before procedures likely to give 1 mSv or more to the uterus. New Zealand (ORS C1): before any procedure that could give a significant fetal dose. Canada (CAMRT): ask all female patients of childbearing age, for example 10 to 55.

Sources: HPA RCE-9 2009; ACR-SPR Pregnancy PP; RCR IR(ME)R 2026; ARPANSA RPS C-5; ORS C1 2024; CAMRT BPG; Health Canada SC35; BIR shielding 2020

What you see

A patient, parent or carer asks for a lead shield, or you are unsure whether your department still uses gonad or fetal shielding.

Why it happens

Diagnostic doses have not been linked to measurable harm to the gonads or fetus, shields do not stop scatter inside the body, and shields can hide anatomy, cause repeats, and raise the exposure if they cover an AEC chamber (AAPM, NCRP). Regional guidance now differs on routine shielding (below).

Try first

1. Follow your local shielding policy, which reflects your regional rules.
2. Where your policy does not shield routinely and a patient asks for a shield, explain that careful collimation and technique protect them more; a shield may then be offered after discussion where it will not interfere with the examination.
3. Never let a shield cover the anatomy of interest or an active AEC chamber.

Stop and escalate when

The patient will not proceed without a shield that would compromise the examination: ask the radiologist or practitioner.

Where sources differ

AAPM (2019) and NCRP (2021): routine gonad and fetal shielding should stop. UK (BIR and others, 2020): routine shielding is not recommended, with narrow exceptions set out in its tables; shielding more than about 5 cm from the beam has a negligible effect. Health Canada (2024): patient shielding is “not recommended unless there are clear benefits”, and routine gonadal shielding should not be used. Australia: ARPANSA’s safety guide RPS 14.1 (2008), which ARPANSA says can still be used alongside its 2019 code, says to shield the gonads, eyes, breast and thyroid whenever feasible, unless the shield would obscure the anatomy needed. New Zealand: ORS C1 does not address patient shielding. ASMIRT’s standards say to use shielding devices “when appropriate”. Follow your local and regional rules.

Sources: AAPM PS 8-A; NCRP 13; BIR shielding 2020; Health Canada SC35; ARPANSA RPS 14.1; ARPANSA RPS C-5; ORS C1 2024; ASMIRT PPS 2018; IAEA RPOP children; ACR DR PP 2022

What you see

A baby or young child needs an X-ray, will not keep still, or the protocol is designed for adults.

Why it happens

Children are more sensitive to radiation and vary widely in size, so technique is matched to their size. Movement is the main challenge. Grids are not normally needed for small children.

Try first

1. Measure the part thickness and use the child-size technique your protocol sets.
2. Avoid a grid for body parts thinner than about 10 to 12 cm (Image Gently); grids are normally needed only for teenagers and larger children.
3. With AEC, make sure the sensors are inside the field and covered by the anatomy; small anatomy can cut the exposure short. The IAEA notes that AEC is often unsuitable for small children, and exposure charts based on thickness are an alternative.
4. Use immobilisation devices first; explain in words the child understands, and involve the parent (see Holding).
5. Collimate to the area needed; image only the projections the radiologist needs.

Stop and escalate when

The child cannot be kept still with these measures: agree a plan with the radiologist and referrer.

Sources: Image Gently DR; IAEA RPOP children; IAEA SSG-46; ACR DR PP 2022; Health Canada SC35; Don 2013

What you see

A child or a patient who cannot keep still needs holding during the exposure, and someone (a parent, an escort or a member of staff) offers to hold.

Why it happens

Immobilisation devices are used first. Staff should not hold patients. A person who holds acts as a carer or comforter: they are informed about the risk, agree, are not pregnant, stay out of the direct beam and wear a protective apron. Their dose is limited by a constraint set under local or national rules.

Try first

1. Try immobilisation devices first.
2. If holding is needed, ask a relative or friend who agrees after being told the risks; check they are not pregnant and are an adult where possible.
3. Give them a protective apron (and gloves if their hands are near the beam), position them out of the direct beam, and collimate tightly.
4. Never hold the detector or cassette by hand during the exposure.

Stop and escalate when

No suitable carer and the patient cannot be immobilised: ask the radiologist or practitioner how to proceed.

Where sources differ

Dose constraints differ: IAEA SSG-46 says 5 mSv per episode is commonly used for adult carers and 1 mSv per episode for a child carer; the Australian code (ARPANSA RPS C-5) uses 1 mSv per diagnostic examination; UK and New Zealand rules leave the constraint to the employer or licence holder. Who may hold: the IAEA, ARPANSA and New Zealand steer away from staff holding; Health Canada allows parents, escorts or other personnel to assist occasionally, with aprons and gloves, but no one should do it regularly; a CAMRT page (2012) allows another healthcare professional. Pregnant carers: the IAEA says carers should not be pregnant and, where one is, the fetal dose is often constrained to 1 mSv per episode; the RCR says a pregnant carer would not normally be designated, but may be with medical physics advice if informed and willing.

Sources: IAEA SSG-46; Health Canada SC35; ORS C1 2024; IR(ME)R 2017; RCR IR(ME)R 2026; ARPANSA RPS C-5; ARPANSA RPS 14.1; CAMRT BPG

What you see

You are taking a mobile X-ray on a ward, in ICU, in the emergency department or in theatre, with other patients, staff and visitors nearby, and the patient has lines, tubes or drains.

Why it happens

The area around a mobile becomes a controlled area during the exposure. Other people are protected mainly by distance, by directing the beam away from them and by aprons for those who must stay. Mobile images are usually of lower quality than department images, and the radiographer is responsible for radiation safety during the exposure.

Try first

1. Ask everyone not needed to move away before the exposure, give a clear warning such as “X-ray”, and wait until they have gone. The IAEA advises asking pregnant staff (other than radiology staff) to leave the vicinity in trauma rooms; apply it to wards as your local rules say.
2. Direct the beam away from other people, even if that means repositioning the patient.
3. Anyone who must stay close (for example with a ventilated patient) wears a protective apron; you wear one too and stand as far away as possible while keeping the patient in view.
4. Before moving the patient or sliding the detector, check lines, tubes and drains with the nursing staff and use transfer aids and enough help.
5. Collimate within the detector and check the image before leaving.

Stop and escalate when

You cannot control the area or keep people out of the beam: do not expose until you can.

Where sources differ

Distance. IAEA SSG-46 says other staff typically at least 3 m from the patient or behind a barrier; Health Canada says the operator at least 3 m from the tube unless wearing protection; ARPANSA’s safety guide RPS 14.1 (2008), still used alongside its 2019 code, says the operator stands at least 2 m from the tube and outside the primary beam, wearing an apron; UK and New Zealand rules give no distance and leave it to local rules. Follow your local rules.

Sources: IAEA SSG-46; HSE L121; CAMRT BPG; Health Canada SC35; ARPANSA RPS 14.1; Brady 2020; Schaefer-Prokop 2008; ARPANSA RPS C-5

What you see

The patient is in isolation or on contact, droplet or airborne precautions, or a detector and mobile unit move between patients.

Why it happens

Detectors and mobile units touch many patients and can carry infection between them. Routine practices apply to every patient, with extra precautions for patients in isolation.

Try first

1. Follow your facility’s infection control procedures: hand hygiene, and the gloves, gown, mask and eye protection the precautions require.
2. Decontaminate the detector, cassettes and positioning aids after each patient following the manufacturer’s instructions, or cover them with a single-use cover (SoR).
3. Isolate electrical equipment from the mains before cleaning it (SoR), and do not spray liquid directly onto a mobile unit: excess liquid has caused a unit to move on its own.
4. Where your procedure uses detector bags, remove the outer layer without contaminating the inner one.

Good to know

The only professional-body guidance found is the SoR’s 2006 guidance on health care associated infections, now in its archive, written for cassettes before DR detectors were common, and not covering isolation-room technique; one hospital’s published pandemic technique (Brady 2020) describes bagging detectors. Your local infection control policy takes precedence.

Sources: CAMRT BPG; SoR HCAI 2006; Brady 2020; FDA recalls

What you see

The patient cannot stand for an erect view, cannot keep a position, or is too unwell or in too much pain to move.

Why it happens

Holding a position takes effort, and pain, weakness and confusion cause movement. Positioning aids and adapted technique let the patient keep still without effort.

Try first

1. Adapt the projection to the patient’s condition (for example seated or supine instead of erect) as your protocol allows, and tell the radiologist which projection was taken.
2. Use positioning and immobilisation aids (pads, sandbags, straps) so the patient does not have to work to keep still.
3. Use transfer aids and enough staff to move the patient; do not lift alone (see Fall).
4. Give the instruction and expose quickly, with the shortest exposure time your technique allows.

Stop and escalate when

The patient deteriorates, faints or cannot be positioned at all: stop, get help, and ask the radiologist what is still needed.

Sources: ASMIRT PPS 2018; Health Canada SC35; IAEA SSG-46; CAMRT BPG

What you see

A trauma patient arrives immobilised, possibly with a spinal injury, in the emergency department or resuscitation room.

Why it happens

In suspected spinal injury, protection of the spine and the choice of imaging follow the trauma guideline. In England, NICE recommends CT for the adult cervical spine when imaging is indicated, and plain X-ray as the first-line test for thoracic or lumbosacral injury without neurological signs.

Try first

1. Keep the immobilisation in place and move the patient only as the trauma team directs, with in-line stabilisation.
2. Confirm the imaging requested matches your trauma protocol.
3. In a shared trauma room: direct the beam away from staff and other patients, ask staff who must stay to wear aprons, ask pregnant staff (other than radiology staff) to leave the vicinity (IAEA), and give a warning before the exposure.
4. For children, NICE allows the peg view to be omitted when the child cannot follow commands or open their mouth.

Stop and escalate when

X-rays are technically inadequate and injury is still suspected: tell the radiologist; CT may be needed.

Good to know

No guideline was found that sets out horizontal-beam or cross-table technique for immobilised patients; follow your local protocol.

Sources: NICE NG41; NICE NG232; IAEA SSG-46

What you see

The patient is large: near the table or stand limit, wider than the detector, or images are noisy or need multiple exposures.

Why it happens

Weight and girth matter more than body mass index. Larger patients attenuate more, so images are noisier and exposure times longer, and the anatomy may be bigger than one detector.

Try first

1. Record weight and girth, and check them against your equipment’s limits; exceeding a manufacturer’s weight limit can damage the equipment.
2. Plan overlapping images when the anatomy is larger than the detector.
3. Position the patient so the thinnest layer of tissue is next to the detector, and raise the kV as your protocol allows; these reduce dose.
4. Use landmarks you can find through tissue, and check the image for noise and motion.

Stop and escalate when

The patient exceeds the equipment’s limits: do not proceed on that equipment; ask the radiologist or referrer about alternatives.

Sources: Uppot 2018; Yanch 2009; Modica 2011

What you see

The patient does not share your language, is Deaf or hard of hearing, or cannot understand or respond to breathing and positioning instructions.

Why it happens

Identification, pregnancy questions, positioning and breathing instructions all depend on two-way communication; instructions should be clear and in a language the patient understands.

Try first

1. Arrange a qualified interpreter, in person or by a remote service.
2. Agree simple signals and demonstrate the position and the breathing instruction before the exposure.
3. Keep the patient in view and within talking distance, and explain they can stop at any time.

Stop and escalate when

Identification, consent or the pregnancy question cannot be completed reliably: ask the radiologist or practitioner before exposing.

Sources: IAEA SSG-46; CAMRT BPG; HDC Code 1996

What you see

You find that the person X-rayed was not the patient on the request, the wrong side, region or examination was imaged, or images were stored under the wrong patient.

Do this now

1. Stop, check on both patients and confirm who each one is, and tell the radiologist or practitioner now.
2. Keep the images and dose record; do not delete anything. Write down what happened and when.
3. Report it through your incident system as your local procedure directs. A medical physicist estimates the dose received.
4. The referrer and the patient (or their representative) are told what happened, under your open disclosure or duty of candour policy.
5. Correct the records through your authorised correction process.
6. Whether the regulator must be notified depends on your country’s criteria; your employer decides after the preliminary investigation. Do not guess thresholds yourself.

Where sources differ

The frameworks agree on investigating, estimating the dose and informing the referrer and patient; the detail differs. UK (IR(ME)R 2017 regulation 8, CQC guidance): the employer investigates and notifies the enforcing authority of exposures significantly greater than intended, with criteria that differ between England and the other nations; the duty to tell the patient covers clinically significant exposures (not defined in the regulations; RCR guidance helps employers decide), alongside the duty of candour. New Zealand (ORS C1): the licence holder notifies the Director of an overexposure as soon as practicable, investigates with dose estimates, and makes sure the referrer and patient are told of any unintended medical exposure; the HDC expects open disclosure usually within 24 hours. Australia (ARPANSA RPS C-5): the operator reports to the Responsible Person and the radiologist, a medical physicist estimates the dose, the radiologist informs the patient, and reports go to the state or territory regulator as it requires. IAEA SSG-46: investigation with a physicist’s dose estimate, the referrer notified, significant events reported to the regulator.

Sources: IAEA SSG-46; IR(ME)R 2017; CQC SAUE 2024; RCR IR(ME)R 2026; ORS C1 2024; HDC Open Disclosure 2019; ARPANSA RPS C-5; Clunie 2014

What you see

An image needs repeating, or more radiation was given than planned: wrong technique or protocol (such as an adult setting on a child), the wrong detector selected, an exposure with no image, a CR plate exposed twice, or an equipment fault.

Do this now

1. Before repeating, check the image really cannot answer the question (zoom, reprocess) and ask the radiologist if unsure.
2. Find and correct the cause first: technique, detector selection, positioning, collimation or equipment.
3. Record the repeat and its reason so it appears in reject analysis.
4. Equipment fault: stop using the equipment until it is put right, record it and report it (see Equipment and system).
5. Report wrong settings and exposures much greater than intended through your incident process; a medical physicist estimates the dose and your employer decides on notification.

Good to know

In the UK, a technical repeat with no procedural, human, systematic or equipment error (for example images spoiled by movement) is not a statutory notification, but events and near misses are still recorded and analysed locally (CQC). An exposure made on a detector that was not selected, or without generating an image, still counts as an exposure.

Sources: CQC SAUE 2024; IR(ME)R 2017; ARPANSA RPS C-5; ORS C1 2024; IAEA SSG-46; AAPM TG-305; Martin 2017

What you see

After an X-ray, it emerges that the patient was pregnant at the time.

Do this now

1. Tell the radiologist or practitioner and report it through your incident process: an inadvertent exposure of an embryo or fetus is investigated.
2. A medical physicist estimates the dose to the embryo or fetus; it needs the examination details and the stage of pregnancy.
3. The referrer and the patient are informed, and counselling is given by a clinician with the right experience, using the estimate.
4. Record how pregnancy status was checked before the exposure.
5. Your regional rules decide whether the regulator must be told.

Good to know

For the counselling clinician, not for the radiographer to give: the HPA table puts chest, skull and thoracic spine radiographs at 0.001 to 0.01 mGy to the fetus, abdomen, pelvis and hip at 0.1 to 1.0 mGy, and lumbar spine at 1.0 to 10 mGy; the ACR-SPR parameter says most radiographic examinations deliver much less than 20 mGy to a conceptus.

Sources: IAEA SSG-46; HPA RCE-9 2009; ACR-SPR Pregnancy PP; CAMRT BPG; ARPANSA RPS C-5; ORS C1 2024; IR(ME)R 2017

What you see

A dosimeter is lost, damaged, left in the X-ray room or exposed when not worn, a dose report is higher than usual, or a member of staff thinks they were in the beam.

Why it happens

Dosimeters estimate each worker’s dose, and investigation levels give early warning that working practice needs review. A dosimeter is worn only by the person it was issued to, in the position your provider specifies, and stored away from radiation when not worn.

Do this now

1. Lost or damaged dosimeter: tell your radiation protection officer or supervisor straight away; the dose for that period is estimated and recorded.
2. Dosimeter left in the room or exposed when not worn: tell the radiation protection officer, so the reading is not wrongly recorded as your dose.
3. A higher than usual reading, or a possible exposure in the beam: tell the radiation protection officer; the investigation looks at your work routine, incidents and colleagues’ doses.
4. Pregnant staff: tell your employer so duties can be reviewed under your local rules.

Where sources differ

Where to wear it with an apron: IAEA SSG-46 says under the apron if you only sometimes wear one, and outside at the collar only if you always wear one, unless national rules say otherwise; Health Canada says under the apron. Follow your dosimetry provider. Investigation levels differ: the UK caps the investigation level at 15 mSv a year and employers usually set it much lower (the UK annual dose limit for adult employees is 20 mSv); IAEA SSG-46 gives examples such as more than 0.5 mSv a month under an apron; New Zealand leaves levels to the licence holder, and Australia to the Responsible Person.

Sources: IAEA SSG-46; HSE L121; ORS C1 2024; ARPANSA RPS C-1; Health Canada SC35; NZ RSA 2016

What you see

A detector is dropped onto a patient, a wall stand or tube moves onto a patient, or a patient is caught by moving equipment.

Why it happens

Such events have been reported to the US FDA, including a detector falling on a patient’s eye, wall stands moving down onto patients’ legs, automatic tube movement towards a seated patient, and portable detectors dropped from tables. Detectors are handled close to patients; one technologist reported a wrist fracture after repeatedly lifting a detector one-handed.

Do this now

1. Stop all movement (use the emergency stop if needed), check the patient for injury and get medical help.
2. Do not move the equipment unless it is still a danger; record what happened, including any error messages.
3. Take the equipment out of use and report it to service and through your incident system.
4. Know where the emergency stops are and how your system’s automatic movements work before using it.

Good to know

These events come from US adverse event reports, which cannot show how often they happen. Hold detectors with two hands and use the holders provided (a general step).

Sources: FDA MAUDE; FDA recalls; Walz-Flannigan 2018

What you see

A mobile X-ray unit hits a person, wall or door, moves on its own, or does not stop.

Why it happens

Staff have been injured by motorised mobile units: feet run over, people pinned against walls, and fingers caught in the tube arm. Manufacturers have recalled units for brakes that did not apply and for uncommanded movement.

Do this now

1. Release the drive handle and press the emergency stop; keep people clear.
2. Check the person struck and get first aid or medical help.
3. Take the unit out of use, label it, and report it to service and through your incident system.
4. Drive slowly, look where you are going, and do not hold a door with one hand while driving with the other (a general step).

Sources: FDA MAUDE; FDA recalls; BfArM 2013

What you see

A patient falls, slips or is lowered to the floor while getting on or off the table, standing at the erect stand, during a transfer, or while walking in the department.

Why it happens

Falls happen in imaging departments, and most patients who fall have at least one risk factor; some are injured. Radiography rooms are among the commonest places for them.

Do this now

1. Call for help, stay with the patient and check whether they are hurt; get a doctor to assess them (a general step; follow your falls procedure).
2. Do not lift the patient alone: use the handling aids and the number of staff your assessment calls for.
3. Record the fall and report it through your incident process, and follow up as your policy requires.
4. Before the next transfer, reassess fall risk: it can change during the examination.

Good to know

Prevention: assess each patient’s fall risk under your policy, help patients onto and off the table and stand, keep the call button within reach, and use slide sheets, slide boards or lifts as needed.

Sources: CAMRT BPG; Abujudeh 2011

What you see

The patient is unresponsive and not breathing normally, or only gasping.

Do this now

1. Call the resuscitation team on your emergency number and stop any exposure.
2. Start chest compressions at once: CPR is started for anyone unresponsive and not breathing normally, even with occasional gasps. Staff trained and willing to give rescue breaths do so.
3. Attach a defibrillator or AED as soon as it arrives and follow its prompts. Minimise pauses in compressions.
4. Move equipment, including a mobile unit, out of the team’s way.
5. Afterwards, record the times and events and report the incident.

Sources: ANZCOR Guideline 8 2021; FDA MAUDE