Diagnostic display QA for radiology: a plain-language quality assurance guide

Why diagnostic display QA matters

Every radiology report rests on an unstated assumption: that the display showed the radiologist everything the image contained. Display quality assurance is how a practice makes that assumption safe.

Displays are physical devices and their performance drifts. Backlights age and lose brightness, uniformity degrades, pixels fail, and a display that was excellent at installation can be quietly mediocre a few years later. Because the change is gradual, nobody notices from day to day, which is exactly why scheduled QA exists: subtle low-contrast findings are the first thing a degraded display stops showing, and the radiologist has no way of knowing what they were not shown.

There is also an accreditation and professional dimension. Practice accreditation and the RANZCR standards of practice for clinical radiology set expectations that displays used for diagnosis are fit for purpose and subject to a quality assurance program. Being able to produce QA records for your display fleet is the difference between asserting your reading environment is adequate and demonstrating it.

Display QA is not exotic or expensive relative to what it protects. It needs someone to own it, a modest amount of tooling, and a calendar. This guide covers the concepts in plain language; for help standing a program up, Trucell’s medical displays and radiology IT support services do this work with imaging practices.

Luminance, calibration, and GSDF in plain language

Three terms carry most of the technical weight in display QA.

Luminance is how much light the display emits, and the two figures that matter most are the maximum (how bright white gets) and the minimum (how dark black stays). The ratio between them sets the contrast range available for the image. Diagnostic displays are specified to reach and hold a much higher maximum luminance than consumer monitors, and the applicable standards set minimum values for diagnostic use. Those values are numbers you should look up, not remember: confirm the current thresholds in the RANZCR standards of practice and the standards they reference.

Calibration is measuring the display with an instrument (a photometer, either built into the display or applied to the front of it) and adjusting the display’s response until it behaves the way the standard says it should. Calibration is not a one-off, because the thing being calibrated keeps changing as the hardware ages.

DICOM GSDF conformance answers the question “behaves in what way?”. The DICOM standard defines the Grayscale Standard Display Function, a curve describing how image values should map to displayed brightness so that grey-level steps are perceptually even for a human observer. When a display is calibrated to the GSDF, a greyscale study looks the way it is supposed to look, and looks the same on the next conformant workstation. Without it, two radiologists can see meaningfully different renderings of the same image and neither would know.

Purpose-built diagnostic displays, including the medical display ranges Trucell supplies through its medical displays solution, are designed for this workflow: high sustained luminance, front sensors or integrated photometers, and vendor QA software that automates calibration and record keeping.

Acceptance testing vs ongoing QA

These are two different activities and a program needs both.

Acceptance testing happens once per display, at installation or after a move or repair. It establishes, with measurements, that this specific display in this specific location meets the applicable specification: luminance range, uniformity, GSDF conformance, absence of pixel defects, and correct configuration of the graphics pipeline driving it. Acceptance testing is your protection at the moment of purchase, and its records are the baseline every later result is compared against. Skipping it means every future QA result floats with no reference point.

Ongoing QA is the scheduled program for the rest of the display’s life. It typically layers three things: continuous automated monitoring where the display fleet and its QA software support it, periodic measured tests that repeat the key acceptance measurements and confirm the display still conforms, and frequent lightweight visual checks where a staff member displays a standard test pattern and confirms the expected elements are visible. The widely used test patterns and test methods come from the AAPM TG18 report and the IEC 62563 series, and the RANZCR standards of practice indicate what is expected in Australian and New Zealand practices.

The other half of ongoing QA is boring and essential: records. Which display, tested when, by whom, with what result, and what was done when something failed. A failed QA result must have a consequence (recalibration, repair, replacement, or removal from diagnostic use), and the record should show it.

Ambient lighting in reading rooms

The display is only half of the viewing system; the room is the other half. The eye adapts to the overall light level around the screen, and image contrast that is visible in a suitably dim room disappears in a bright one. Reflections and glare from windows or poorly placed lights add a veil of light over the screen surface that washes out exactly the low-contrast detail diagnostic displays are built to show.

Practically, that means reading rooms should have controlled, consistent, reasonably low lighting, no light sources or bright windows behind the radiologist or reflected in the screen, and ideally indirect lighting that can be set to a repeatable level. It also means ambient light is part of QA, not separate from it: the applicable standards treat ambient luminance as a measured parameter, displays are calibrated for an assumed ambient level, and a room that is much brighter than that assumption quietly invalidates the calibration. Confirm the recommended ambient conditions and how to account for them in the RANZCR standards of practice and the IEC 62563 series rather than working from rules of thumb.

Two habits cost nothing and help: keep reading-room lighting the same during QA testing as during reporting, and re-check any room where lighting changes (new fitout, moved desks, a new window treatment) even if the displays themselves have not moved.

When clinical review displays differ from diagnostic displays

Not every screen in an imaging practice needs to be a diagnostic display, and pretending otherwise makes QA programs collapse under their own weight. The useful distinction is by task:

• Diagnostic (primary interpretation) displays are the screens radiologists report from. These carry the full specification: diagnostic-grade hardware, acceptance testing, calibration to the GSDF, scheduled QA, and a controlled reading environment. Mammography reporting carries its own, stricter display expectations within this category; confirm the current requirements in the applicable standards and accreditation programs before specifying mammography workstations.
• Clinical review displays are everything else that shows images to a clinician: referrer viewing, ward and theatre screens, consultation-room monitors. They are not used to make the primary read, so they carry a lighter expectation: reasonable quality, sensible configuration, and periodic checks, but not the full diagnostic specification.

The risk sits at the boundary. Problems arise when the line is undefined and primary interpretation quietly happens on review-grade screens: an urgent read done at a ward PC, or a home report from a consumer monitor. A practice should be able to say exactly which displays are approved for primary reporting, and its PACS workflow should make the approved workstations the natural place that work happens. This is as much a workflow and PACS and RIS configuration question as a hardware one.

A practical QA cadence

The correct frequencies and pass criteria come from the current RANZCR standards of practice and the standards they reference, and they are revised over time, so treat the following as a structure to fill in from those documents rather than a set of numbers:

1. At installation, relocation, or repair: acceptance test. Full measured test against specification, results recorded as the display’s baseline.
2. Continuously, where supported: automated monitoring. Modern diagnostic displays with integrated sensors can self-check and report into fleet QA software. Automation catches drift between scheduled tests, but it supplements the scheduled program rather than replacing it.
3. Frequently: quick visual check. A staff member displays a standard test pattern (the AAPM TG18 patterns are the common choice) and confirms the expected low-contrast elements are visible and the screen is free of new defects. Minutes per workstation.
4. Periodically: measured QA test. Repeat of the key acceptance measurements (luminance, uniformity, GSDF conformance) with a calibrated instrument, compared against the baseline, with recalibration where needed.
5. On failure: a defined response. Recalibrate, repair, replace, or remove the display from diagnostic use, and record the action taken.
6. Annually or on change: review the program. New displays, moved rooms, lighting changes, and updates to the RANZCR standards all trigger a review of what is being tested and against what criteria.

Assign each layer an owner by name. Programs fail from ownership gaps far more often than from technical difficulty, and a fleet QA platform only helps if someone is responsible for acting on what it reports.

Standards and guidance to work from

For an Australian or New Zealand practice, the reference points to hold your program against are, by name:

• The current RANZCR standards of practice for clinical radiology, which set the professional expectations for image display, reading environments, and teleradiology arrangements in this region. Where a specific threshold or test frequency matters, confirm it in the current edition rather than in summaries like this one.
• The IEC 62563 series on medical image display system evaluation, which defines test methods for assessing display performance.
• The AAPM TG18 report, the widely used American Association of Physicists in Medicine reference for display assessment, whose test patterns and methods underpin much day-to-day display QA.
• The DICOM standard’s Grayscale Standard Display Function, which defines the greyscale response that calibration targets.
• Display and QA software documentation from your display vendor, which translates the standards into the concrete procedure for your specific fleet.

These documents are updated over time, and thresholds or recommended intervals you memorised from an earlier edition may no longer be current. When a number matters, look it up in the current edition or ask your medical physics or IT partner to confirm it.

A note on scope

This is general guidance, not compliance, accreditation, or medical physics advice. Display QA requirements, recommended test frequencies, luminance thresholds, and mammography-specific expectations are set by the current editions of the RANZCR standards of practice, the relevant IEC and AAPM publications, and the applicable accreditation programs, and they change over time. Confirm current requirements against those documents and with your accreditation body, a qualified medical physicist, or a suitably qualified adviser before relying on any of the above.