“Density” and “contrast” remain common teaching terms, but they must be used differently for film and digital imaging. Optical density is the blackening of processed film. In digital radiography, exposure, detector signal, displayed brightness, contrast and noise are related but separate properties. kVp and mAs shape the acquired data; processing then shapes its appearance.
Image quality is task dependent. A useful introductory framework includes exposure/noise, contrast, spatial resolution and geometric fidelity, while recognizing that patient positioning, anatomy, scatter, detector performance and processing interact.
Film: optical blackening. Digital: detector exposure and signal—not displayed brightness. At fixed conditions, mAs changes detector exposure approximately proportionally.
Signal or brightness differences between regions. Subject attenuation, beam energy and scatter affect acquired contrast; processing strongly affects displayed contrast.
Ability to represent small structures. Detector sampling and response, focal spot, geometry and motion all matter; exposure can affect whether detail is detectable through noise.
Size or shape misrepresentation caused by projection geometry. Source, object and detector distance and alignment all matter.
For screen–film imaging, optical density is a logarithmic measure of film blackening and depends on receptor exposure and the film characteristic curve. For CR/DR, the raw detector response changes with incident air kerma, but the presentation image is rescaled and processed. An underexposed digital image may have normal brightness yet excessive quantum noise; an overexposed image may also look normal while carrying avoidable patient dose. Only extreme detector exposure may saturate or clip data.
At fixed kVp, filtration and geometry, mAs (milliampere-seconds) is approximately proportional to X-ray photon fluence, tube output and detector air kerma:
This does not mean double digital brightness. The effects below refer to detector exposure when other conditions are unchanged:
| Variable | Change | Effect on detector exposure / signal |
|---|---|---|
| mAs | 2× | Approximately 2× at fixed kVp and geometry |
| kVp | Increase | Usually increases tube output and transmission; magnitude is system, spectrum and patient dependent |
| Source-to-image distance | 2× | Approximately 0.25× incident fluence in air (inverse-square relationship) |
| Added filtration | Increase | Removes more low-energy photons and hardens the beam; net detector exposure depends on technique and patient |
| Patient thickness | Increase | Less primary transmission and generally more scatter |
| Grid | Add or increase selectivity | Reduces total detector exposure; compensation depends on measured/manufacturer grid performance, not ratio alone |
Within generator, tube and detector operating limits, combinations that yield the same mAs produce approximately the same photon fluence at fixed kVp:
A shorter time can reduce motion blur, provided the selected mA and focal spot are permitted and the generator remains accurate. Reciprocity is not a guarantee across equipment limits or every receptor.
Assess technique with the image, the standardized exposure index (EI) and deviation index (DI), and local reject/quality criteria—not brightness alone. EI estimates detector response under defined calibration, while DI compares EI with a target EI for that examination. EI and DI are not measures of patient dose. Segmentation, collimation, prostheses and processing errors can make an index misleading.
Contrast is a difference in signal or displayed brightness between regions. Detectability also depends on noise, object size, spatial resolution and the viewing conditions—not contrast alone.
Determined by the patient's anatomy and the X-ray beam energy. Different tissues attenuate X-rays differently based on:
In screen–film radiography, the film characteristic curve links exposure to optical density and limits useful latitude. In digital radiography, detector response, bit depth and dynamic range determine what is captured; processing, lookup tables and window/level determine presentation contrast. Processing can redistribute visible grayscale but cannot recover a signal buried in quantum noise, lost to saturation, blurred by motion, or degraded by scatter.
kVp changes maximum photon energy, the spectrum and tube output. For a given object, higher kVp generally increases penetration and reduces subject contrast. It does not by itself dictate the displayed contrast of a processed digital image:
Keep acquisition and display separate: kVp changes the X-ray information arriving at the detector; processing changes how that information is displayed. A display adjustment may make an image look more contrasty without improving its acquired contrast-to-noise ratio.
The traditional 15% rule is an approximate screen–film-era technique heuristic:
A roughly 15% increase in kVp may approximately double receptor exposure for some general-radiography conditions.
As a starting estimate, one may halve mAs after a 15% kVp increase, or double mAs after a 15% decrease, to seek similar detector exposure. But the relationship varies with generator waveform, filtration, patient thickness/composition, detector energy response and selected kVp. Verify it using an established technique chart, EI/DI trends and local physics/quality-assurance data.
For a teaching calculation only, starting from 70 kVp at 10 mAs:
In general projection radiography, most patient-generated scatter that reaches the detector arises from Compton interactions. A scattered photon changes direction and loses some energy; if detected, it contributes signal in a location that does not faithfully represent the primary-beam attenuation path.
Scatter adds a broad, spatially varying background and quantum noise. It reduces subject contrast by a factor related to the scatter-to-primary ratio and can reduce contrast-to-noise ratio. It is not necessarily uniform across the field.
| Factor | Effect |
|---|---|
| Larger field size | More irradiated volume generally raises the scatter fraction; collimate to the anatomy required |
| Greater patient/part thickness | Scatter fraction generally increases, tending toward saturation at large thickness |
| Beam energy | Interaction probabilities and escape direction change; detector scatter fraction may have only weak energy dependence for efficient receptors |
| Composition and geometry | Photoelectric and Compton contributions, attenuation and the scatter reaching the detector all vary; atomic number alone is not a useful clinical scatter rule |
An antiscatter grid contains aligned radiopaque strips separated by low-attenuation interspaces between the patient and receptor. It preferentially rejects obliquely incident scatter, but it also removes some primary radiation. Maintaining detector exposure therefore requires increased technique and usually increased patient dose. Grid use must be justified by the improvement for the imaging task.
One geometric descriptor is the grid ratio:
Grid ratio = lead-strip height / interspace width
| Grid concept | Accurate interpretation |
|---|---|
| Grid ratio | Lead-strip height divided by interspace width; a higher ratio generally rejects more scatter but demands more precise centring, alignment and focal distance |
| Bucky/grid factor | mAs with grid divided by mAs without grid for comparable receptor exposure; it depends on the complete grid, beam quality, field and phantom/patient—not grid ratio alone |
| Focused grid | Must be used within its specified focal range and with correct centring and orientation to avoid primary cut-off |
| Alternatives | Tight collimation is always foundational; compression or an air gap may reduce scatter when appropriate to the examination |
A technique chart is an equipment-, detector- and examination-specific starting protocol. It should incorporate projection, patient size or measured part thickness, diagnostic task, source-to-image distance, filtration, grid status, detector characteristics, and either manual mAs or appropriately configured AEC.
| Problem | Likely Cause | Fix |
|---|---|---|
| Excessive quantum noise / low CNR | Low detector exposure, attenuation, grid/collimation or processing/segmentation issue | Check positioning, collimation, EI and DI against the exam target; adjust the validated technique rather than automatically doubling mAs |
| Unexpectedly high EI/positive DI | Excess detector exposure or index/segmentation error | Review raw image coverage, collimation and artifacts; if valid, optimize future technique—do not repeat solely to correct EI/DI |
| Flat appearance / poor CNR | Scatter, processing, excessive penetration, or low acquired signal | Confirm protocol/processing, collimate, and evaluate kVp and grid use for the task; windowing cannot restore lost CNR |
| Saturation or clipped anatomy | Extreme exposure or processing failure | Follow local repeat criteria and correct the causal technique/processing issue |
| Unsharp edges | Motion, geometric blur or detector resolution | Correct immobilization/positioning and geometry; shorten time or use an appropriate focal spot within tube limits |
| Grid cut-off | Off-level, off-centre, reversed or out-of-focus grid | Correct grid orientation, centring, perpendicularity and specified focal distance |
For systems conforming to IEC terminology, EI is proportional to the calibrated detector exposure estimate. The target exposure index (EIT) is selected for a particular examination and view. The deviation index is DI = 10 log10(EI/EIT): DI 0 means EI equals the target; +1 is about 26% above target, −1 is about 21% below, +3 is about twice target and −3 about half. Facility-specific acceptable ranges must be based on the task and detector. DI is feedback, not an automatic repeat instruction and not patient dose.
For more detail on how these principles interact, read X-Ray Physics Made Simple: kVp, mAs, Density, and Contrast. For the complete story on how X-rays are generated in the first place, see X-Ray Production: Bremsstrahlung & Characteristic Radiation.
Try these multiple choice questions based on this article. Click an option to check your answer — correct answers turn green, wrong ones turn red.