11.1 Evaluating Image Quality

What Image Quality Means

Image quality describes how faithfully a radiograph represents the patient's anatomy. ARRT's Image Production domain (51 scored items, 25.5% of the exam) expects you to name the quality factors, state what controls each one, and critique a finished image. The six factors you must master are brightness, contrast, spatial resolution (recorded detail), noise (quantum mottle), distortion, and the exposure indicator (EI). Chapter 9 derived the exposure factors and Chapter 10 covered digital hardware; here we focus on judging the finished image.

Brightness and Contrast

Brightness is the amount of luminance in the displayed image; the analog film equivalent is density (degree of blackening). In digital imaging, brightness is set by the look-up table (LUT) during processing and can be re-adjusted with window level, so the receptor exposure no longer controls displayed brightness the way mAs controlled film density. Nevertheless, milliampere-seconds (mAs) remains the primary controller of the receptor exposure that feeds the histogram. Too little exposure yields a noisy image; too much wastes dose.

Contrast is the difference in brightness between adjacent structures — it is what makes anatomy distinguishable. Contrast is described by scale. Long-scale (low) contrast shows many shades of gray with small differences between them and is produced by high kVp. Short-scale (high) contrast shows few, boldly different tones and is produced by low kVp. Kilovoltage (kVp) is the primary controller of contrast scale because it governs beam penetration and the amount of scatter produced. Scatter radiation adds unwanted gray fog that lowers contrast; grids and collimation (Sections 11.2 and 11.3) exist to remove it. In digital systems, the processing algorithm rescales contrast, but you cannot recover information the receptor never captured.

Spatial Resolution and Noise

Spatial resolution (also called recorded detail or sharpness) is the ability to image small, closely spaced structures. It is measured in line pairs per millimeter (lp/mm) — the more line pairs resolved, the sharper the image. Screen-film systems reached roughly 5-10 lp/mm; digital detectors are typically limited to about 2.5-5 lp/mm by the size of the detector element (DEL) and pixel. Geometrically, resolution improves with a small focal spot, short object-to-image-receptor distance (OID), long source-to-image-receptor distance (SID), and elimination of motion. A large focal spot enlarges the blur margin (penumbra), degrading detail.

Noise is random graininess that obscures detail. Its dominant form is quantum mottle — the grainy appearance caused when too few x-ray photons reach the receptor. Because image formation is a photon-counting process, insufficient mAs produces poor signal-to-noise ratio (SNR) and visible mottle. The correction is more mAs (more photons), never more digital processing, which only amplifies existing noise. This is a favorite exam trap: an underexposed, mottled digital image should be repeated with increased mAs, not brightened at the workstation.

Distortion and the Exposure Indicator

Distortion is misrepresentation of the true size or shape of anatomy. Size distortion (magnification) always enlarges and results from increased OID or decreased SID; it is minimized by keeping the part close to the receptor and using a long SID. Shape distortion — elongation or foreshortening — results from misalignment of the tube, part, or IR planes and from improper central-ray angulation.

The exposure indicator (EI) is the manufacturer's number reporting the radiation exposure received by the image receptor. To standardize brands, the IEC/AAPM adopted the deviation index (DI), computed as DI = 10 x log10(EI / EIT), where EIT is the target exposure. A DI of 0 means receptor exposure matched the target. Each +1 is about a 26% overexposure; +3 is roughly double the target (dose creep), and -3 is roughly half (underexposure with mottle). Most departments accept DI within about +/-1 and require repeat evaluation beyond about +/-3. A persistently negative DI (for example -3) across a caseload warns of chronic underexposure and mottle, while a persistently positive DI warns of dose creep — technologists reflexively overexposing because digital rescaling hides the extra dose. Because the display normalizes brightness, the EI/DI is your only objective read on whether the patient received the right dose.

Exposure Latitude and Dynamic Range

Digital receptors have wide exposure latitude (dynamic range) — the span of exposures that still yield a usable image after rescaling. This tolerance is a double-edged sword: it forgives small technique errors but conceals both underexposure (until mottle appears) and overexposure (until the EI/DI flags it). The lesson for image critique is that a digital image can look correctly exposed while still being seriously under- or over-dosed, so brightness alone can never confirm proper technique.

Systematic Image Critique

Critique every radiograph the same way so nothing is missed: (1) anatomy — is the required structure fully demonstrated? (2) positioning/rotation — are symmetric landmarks equal? (3) exposure/brightness and EI/DI — is the receptor exposure appropriate? (4) contrast — is scale suitable for the part? (5) spatial resolution — any motion or geometric blur? (6) collimation, markers, and artifacts — is the field limited, is the anatomical side marker present, and are there processing or hardware artifacts? A common exam trap presents a mottled but adequately bright digital image and asks for the fix: the answer is to repeat with increased mAs, because judging a digital image by brightness alone is wrong. Displayed brightness is normalized, so you must read the EI/DI together with visual receptor exposure to know whether dose was appropriate before accepting or repeating the study.