Quality assurance

Quality assurance is a requirement of IRR 1999 and each hospital should establish its own quality manual detailing:

  • What tests have to be done
  • How the tests should be done
  • How often the tests should be done
  • How the test results are recorded and analysed
  • What the acceptable margin of deviation from the standard is
    • If test differs by a margin that requires action to rectify it = remedial level
    • If test differs by a substantial margin that means equipment no longer fit to use = suspension level
  • What to do if a test is failed

Institute of Physics and Engineering in Medicine (IPEM) report 91: Recommended standards for the routine performance testing of diagnostic x-ray systems.

IPEM report 91 divides QA tests into two levels, which it calls level A and level B.

Level A tests:

  • Are generally quick and simple
  • Do not need expensive or complex equipment
  • Do not need detailed analysis
  • Are done frequently
  • Are usually done by the equipment user

Level B tests:

  • Take longer
  • Might require expensive or complex equipment
  • Need more analysis
  • Might be done less frequently than level A tests
  • Are often done by medical physics departments or by manufacturers’ engineers


Out now in paperback and on Kindle
Written by radiologists, for radiologists with plenty of easy-to-follow diagrams to explain complicated concepts. An excellent resource for radiology physics revision.


FRCR Physics Notes: Medical imaging physics for the First FRCR examination

Note: Not available in all regions.

X-ray set tests

X-ray tube output

  • Tested every 1-2 months
  • Ionisation chamber placed at a known distance from the x-ray tube. The measurements of dose (using an electrometer) are made for various exposures to determine:
    • Dose per mAs (output) for range of exposures
    • Whether output varies with mA
    • How output varies with kV
    • Repeatability - Whether output is consistent when same exposure repeated
      • Remedial level = ± 10%
      • Suspension level = ± 20%
    • Consistency - whether output has changed since the baseline set of QA checks
      • Remedial level = ± 20%
      • Suspension level = ± 50%

X-ray tube kV

  • Tested every 1-2 years
  • Potential measured using an electronic kV meter. The kV is then measured at a range of different exposure settings.
    • Is it accurate - do we get the value we have selected?
    • Does it change if we change the mA range or exposure time?
    • Does it vary during the exposure?
  • Remedial level = ± 5% or ± 5 kV from baseline, whichever is greater
  • Suspension level = ± 10% or ± 10 kV from baseline


  • N.B. not regularly measured as doesn't change.
  • Dose for fixed exposure measured with varying thicknesses of aluminium placed in the beam.
  • The thickness that gives 50% transmission (half original dose) is the half-value thickness (HVT) or half value layer (HVL).
  • Data is available that enables the filtration to be estimated from the HVT.

Automatic exposure control

  • Some tests are done for level A and some for level B.
  • The AEC terminates the exposure once the film has received an appropriate level of dose.
  • It should produce a consistent optical density in the film for a wide range of tube potential (kV) and for a wide range of patient thickness.
  • Perspex or water blocks used to simulate a patient
  • A series of exposures at different tube voltages and using a different thickness of material is taken.
  • The mAs and the detector dose indicator measurement is recorded.
  • AEC sensitivity
    • Tested every 1-3 months
    • A 1 mm copper in the beam is imaged with exposure under the AEC device control then the whole detector is irradiated. The mAs and detector dose indicator (DDI) reading is then recorded.
    • Remedial level = baseline ± 25%
    • Suspension level = baseline ± 50%
  • Guard timer operation
    • Tested annually
    • The guard timer terminates the exposure after a certain amount of time
    • It is tested by using a low kV exposure with lead blocking the AEC chambers and ensuring the exposure is terminated at the guard timer setting.
  • AEC consistency
    • Tested annually
    • The DDI and mAs is checked between the AEC chambers to ensure consistency between them.
    • Remedial level = baseline ± 30%, mean ± 20%
  • AEC repeatability
    • Tested annually
    • The mAs and DDI of successive repeated exposures is measured using the same AEC settings
    • Remedial level = mean ± 20%
  • AEC reproducibility
    • Tested annually
    • Testing is similar to the AEC consistency tests but a larger range of kV and thickness of phantoms is used.
    • Remedial level = baseline ± 30%
    • Suspension level = baseline ± 60%

Light beam alignment

  • Tested every 1-2 months.
  • Edges of light beam marked on film and film exposed. Area of exposed field then compared to light field
  • Remedial level = 1 cm misalignment on any side at 1 m from focal spot
  • Suspension level = 3 cm

Focal spot measurement

N.B. most physics services do not measure this regularly as it doesn't change and any faults can be picked up from the image quality and tube output tests.

  • Pinhole
    • Pinhole a few microns across (smaller than the focal spot) radiographed.
    • Taking into account the distance from the focus to the pinhole and the focus to film distance you can then calculate the size of the focal spot by measuring the image produced on the film.
    • Can estimate size, shape and irregularities in the focal spot with this method
  • Star test object
    • An array of radiating lead spokes is imaged with a geometric magnification of approximately 3
    • The spokes at the centre will come to a point at which they can no longer be distinguished by the system due to being too small and close together.
    • The diameter of the blurred area can be used to calculate the focal spot size

CR and DR radiography

There are a few tests specific to computed and digital radiography as outlined below.

Detector dose indicator (DDI)

  • The detector dose indicator measures the dose received at the detector. There are several tests performed to guarantee the accurate functioning of the DDI.
  • DDI repeatability and reproducibility
    • Tested annually
    • Remedial level = baseline ± 10%
    • Suspension level = baseline ± 20%

Image quality

  • Low contrast sensitivity
    • Tested every 4-6 months
    • The Leeds Test Objects Ltd is used to ensure the system is still able to image low contrast items
    • Remedial level = baseline ± 2 groups
  • Threshold contrast detail detectability
    • Measured annually
    • A test object with an appropriate filter and kV is imaged and the contrast that can be accurately imaged is measured.
  • Limiting spatial resolution
    • Tested every 4-6 months
    • A lead grating resolution bar pattern is used to assess the highest spatial resolution the system can image accurately
    • Remedial level = baseline minus 25%
  • Uniformity of resolution
    • Tested annually
    • A fine wire mesh is imaged and checked for blurred areas and discontinuities
    • Remedial level = increase in blurring from baseline
  • Measured uniformity
    • Tested annually
    • An image is obtained with no object in the field. An ROI is then placed over each quadrant and in the centre. The 5 values are used to calculate the standard deviation divided by the mean value.
    • Remedial level = mean ± 5%
  • Dark noise
    • Tested annually
    • An image is obtained without exposure or with very low exposure. This tests for noise in the system.
    • Remedial level = baseline ± 50%
  • Scaling errors
    • Tested annually
    • A grid and an attenuating object of known dimensions or a lead ruler are used to ensure that the scale of the image is correct.
    • Remedial level = > 2% difference

Σ  Summary
  • Quality assurance is a requirement of IRR 1999 but exact schedule and test list is not specified - up to individual hospital
  • IPEM report 91 provides guidelines
  • Remedial level = action required to improve equipment performance
  • Suspension level = equipment should not be used anymore. Not every piece of equipment has a suspension level

Example testing timeline and summary

(scroll sideways to view whole table)

Equipment tested Frequency of testing Method of testing Performance criteria
X-ray tube output 1-2 months Dose at various exposures measured with ionisation chamber at known distance


  • Remedial = ± 10%
  • Suspension = ± 20%


  • Remedial = ± 20%
  • Suspension = ± 50%
Light beam alignment 1-2 months Light beam field compared to exposed field on film

Remedial = 1 cm misalignment on any side at 1 m from focal spot

Suspension = 3 cm

Automatic exposure control (AEC) sensitivity 1-3 months 1mm copper imaged with exposure under AEC device control. mAs and DDI reading recorded.

Remedial = baseline ± 25%

Suspension = baseline ± 50%

Low contrast sensitivity 4-6 months Uses Leeds Test Object Ltd test object Remedial = baseline ± 2 groups
DR/CR limiting spatial resolution 4-6 months Uses lead grating resolution bar pattern Remedial = baseline minus 25%
AEC ionisation chamber consistency Annual Expose ionisation chambers separately and compare (usually three)

Remedial = baseline ± 30%, mean ± 20%

AEC repeatability Annual mAs and DDI of successive repeated exposures measured Remedial level = mean ± 20%
AEC reproducibility Annual Similar to AEC consistency but larger range of kV and thickness of phantoms used

Remedial level = baseline ± 30%

Suspension level = baseline ± 60%

Focal spot Annual


Pinhole radiographed and size, shape and inconsistencies of focal spot calculated from image produced

Star test object

Radiating lead spokes radiographed and central blurred area size used to calculate focal spot size

Filtration Annual

Half value thickness / layer of filter determined and compared to available data to calculate filtration

Should be equivalent to 2.5 mm aluminium

Detector dose indicator repeatability and reproducibility Annual  

Remedial = baseline ± 10%

Suspension = baseline ± 20%

Threshold contrast detail detectability Annual Test object with appropriate filter and kV imaged  
Uniformity of resolution Annual Fine wire mesh imaged Remedial = increase in blurring from baseline
Scaling errors Annual Grid imaged with object of known length Remedial = >2% deviation from object
Dark noise Annual No exposure or low exposure Remedial = baseline ± 50%
Measured uniformity (DR/CR) Annual Image obtained with no object. ROI over each quadrant and centre. Mean and standard deviation calculated Remedial level = mean ± 5%
/strong> X-ray tube kV 1-2 years Electronic kV meter measures kV at different exposure settings

Remedial = ±5% or ±5 kV from baseline

Suspension = ± 10% or ± 10kV from baseline

Next page: Mammography

  Send us your feedback


Buy FRCR Physics Notes in paperback or on Kindle and follow us for the latest developments and updates via social media.

Buy online
Available on Amazon


Follow on Facebook


Join 10k+ subscribers

Please note: Your email address will never be shared with any 3rd parties. It will only be used for Radiology Cafe communications. Emails are sent less than once a month on average. Read our Privacy policy for more details.