Image Quality
The properties
Spatial resolution is the ability to resolve small high-contrast detail, measured in line pairs per millimetre or by MTF. Contrast resolution is the ability to distinguish low-contrast structures and is limited by noise. Noise includes quantum mottle, electronic noise and anatomical noise. Artefact is the fourth property. Sharpness is degraded by focal spot size, motion, receptor blur and geometry.
Trade-offs
Geometric unsharpness = focal spot size × OID / SOD, so reduce OID and use a small focal spot for sharpness. Magnification = SID / SOD. Grids remove scatter and improve contrast but demand more mAs and exact centring, otherwise grid cut-off appears. Increasing kVp widens latitude and lowers contrast. In digital imaging, window width sets displayed contrast and window level sets brightness; exposure errors appear as noise or saturation rather than as a uniformly dark or light film.
Measurement
Signal-to-noise and contrast-to-noise ratios describe detectability. DQE combines noise and contrast performance. The Nyquist theorem requires sampling at least twice the highest spatial frequency present, otherwise aliasing occurs.
Traps
Confusing spatial with contrast resolution; assuming digital removes the need for correct exposure (exposure creep); confusing magnification with distortion, which is a change of shape.
Worked example
A 1.2 mm focal spot with 10 cm OID and 170 cm SOD gives unsharpness of 1.2 × 10 / 170, about 0.07 mm. Moving the object to 20 cm OID gives about 0.15 mm and magnification of 180/160, about 1.13 - visible blur and enlargement.