Scatter Removal Grids

The antiscatter grid plays an important role for enhancing image quality in projection radiography by transmitting a majority of primary radiation and selectively rejecting scattered radiation. This device is comprised of a series of thin lead strips separated by radiolucent interspaces in a form-factor that matches the detector size. Most grids have a linear geometry in one direction (usually along the long axis of the detector). Parallel grids have lead strips that are focused to infinity (i.e. the primary x-rays have a parallel trajectory). Focused grids have lead strips that are oriented parallel at the center (along the x-ray central axis) and progressively slanted to the periphery to match the beam divergence from the focal spot .to the detector at a specific source to detector distance.

The anti-scatter grid is typically manufactured with lead strips oriented along one dimension separated by a low attenuating interspace material such as carbon fiber or aluminum. For specialized applications, there are cross-hatched grids (lead strips in both directions, perpendicular to each other) for specialized applications such as dedicated chest imaging, and in mammography where a "cellular" grid design made of copper with air interspaces is used clinically by one manufacturer. By selectively allowing primary x-rays to be transmitted and scattered x-rays to be absorbed in the grid, image contrast is significantly enhanced; however, the grid attenuates some of the desired primary x-rays that are incident directly on the lead strips and allows transmission of some scattered radiation photons that have a small scattering angle, or scatter in a direction parallel to the lead strips, or are multiply scattered with an exit angle from the patient that can be transmitted through the grid.

Grids are chiefly characterized by the grid ratio, grid frequency, and focal distance. The grid ratio is a measure of the height of the lead strip to the interspace distance, and is a good measure of the selectivity of primary to scatter transmission. In general, a grid with a higher grid ratio will reject scatter better than a lower grid ratio, due to the limited angle that is allowed by the grid structure. However, a higher ratio grid typically has a higher dose penalty for its use (for screen-film imaging this is known as the "Bucky Factor" which represents the increased dose to the patient when using a grid compared to not using a grid when the film optical density is matched). With digital imaging, there is also a dose penalty when using a grid is used, and the benchmark is the signal to noise ratio (as opposed to film optical density). The grid frequency is a measure of the number of grid lines per unit distance (inches or centimeters), and is in the range of 40 - 50 lines/cm (100-120 lines/inch) for low frequency grids, 50-60 lines/cm (120 - 150 lines/inch) for medium frequency grids, and 60 - 70+ lines/cm (150-170+ lines/inch). Low frequency grids are used with systems having a moving grid assembly (known as a Bucky device) that oscillates during the exposure to blur the grid lines. Medium and high frequency grids are typically used with stationary grid holders (e.g., portable radiography and many digital radiography systems). High frequency grid use is particularly important for digital radiography systems to avoid aliasing artifacts (see section on radiography artifacts) that arise from insufficient sampling of high frequency patterns that are interpreted in the output signal as low frequency (aliased) signals. The grid focal distance is determined by the angle of the lead strip geometry that is progressively increased from the center of the grid to the periphery, to account for the diverging primary x-ray beam emanating from the focal spot. Typical focal distances are 100 cm (40 inches) and 180 cm (72 inches), although there are many specialized grid focal distances. Focal range is an indicator of the flexibility of grid positioning distance from the focal spot, and is a function of the grid ratio and frequency. General purpose grids for portable radiography have a fairly large range (e.g., 80 to 130 cm) while special purpose grids have a much narrower focal range. Grid artifacts arise from improper positioning of the grid device, such as tilting the grid at a non-perpendicular direction to the incident x-ray beam, not centering the grid to the x-ray beam central axis, using a focused grid outside the specified focal range, and placing the grid upside down (converging geometry is directed opposite of the focal spot).

The two images of the AP projection of the knee phantom were obtained at 60 kV at the table top (left) and using the scatter removal grid (Bucky) (right).  The final S numbers of both images were ~350, indicating an air kerma incident on the computed radiography imaging plate of ~6 uGy (0.6 mR) in both cases. The table top image on the left, however, required a technique of 3 mAs whereas the one on the right required 10 mAs, since the scatter removal grid removes most of the scattered photons that emerge from the phantom. The Bucky factor is thus 3.3 (i.e., 10 mAs/3 mAs), and this is a quantitative measure of the increase in patient dose resulting from the use of the scatter removal grid. Note the improvement in image quality achieved by removal of most of the scatter radiation.

The two lateral projection images of the skull phantom were obtained at 75 kV at the table top (left) and using the scatter removal grid (i.e., Bucky) (right).  The final S numbers of both images were ~100, indicating an air kerma incident on the computed radiography imaging plate of ~20 uGy (2 mR) in both cases. The table top image on the left, however, required a technique of 4 mAs whereas the one on the right required 20 mAs, since the scatter removal grid removes most of the scattered photons that emerge from the phantom. The Bucky factor is thus 5 (i.e., 20 mAs/4 mAs), and this is a quantitative measure of the increase in patient dose resulting from the use of the scatter removal grid. Note that there is a dramatic increase of image quality achieved by removal of most of the scatter radiation, and well worth the “cost” in additional radiation dose to the patient.

The two AP projection images of the pelvis phantom were obtained at 75 kV at the table top (left) and using the scatter removal grid (Bucky) (right).  The final S numbers of both images were ~240, indicating an air kerma incident on the computed radiography imaging plate of ~8 uGy (0.8 mR) in both cases. The table top image on the left, however, required a technique of 3 mAs whereas the one on the right required 25 mAs, since the scatter removal grid removes most of the scattered photons that emerge from the phantom. The Bucky factor is thus 8 (i.e., 25 mAs/3 mAs), and this is a quantitative measure of the increase in patient dose resulting from the use of the scatter removal grid. Note that there is a dramatic increase of image quality achieved by removal of most of the scatter radiation, and well worth the “cost” in additional radiation dose to the patient.

Also note that the Bucky factor for the abdomen is substantially higher than those for the knee and the skull radiographs; The reason for this is that scatter is reduced with decreasing kV, as well as when imaging predominantly bony structures where most interactions are through the photoelectric effect (Compton scatter dominates for radiographs of soft tissue structures).