|
| |
Antiscatter Grids - An Essential Component of Image Quality - Part 1
by Dr. Frank Ranallo, PhD
A grid may only comprise about 1% of the cost of an imaging system, but its
importance to system performance is much greater than its cost might indicate.
Properly selected and installed high quality grids can dramatically improve
image quality and even decrease patient dose. However grids are often woefully
neglected both in their initial specification and in the testing of their
performance.
In part 2 of this article I will discuss some of the important tests that should
be performed with grids to verify their quality and proper alignment. In this
article I will discuss some of the properties of grids and some of the most
common examples of how things can go wrong in grid selection and use.
Grids are made up of regularly spaced thin septa of lead separated by
interspaces of aluminum or fiber. An 18 by 18 inch grid will contain many
hundreds of lead septa, each 18 inches long but less than 1 cm high. The first
important properly of a grid is the orientation or direction of the grid lines
(the direction of the long dimension of the grid strips). This is important to
know since one is allowed to image by angling the x-ray tube along the direction
of the grid lines, but not by angling in any other direction. The direction of
the grid lines is generally indicated by a line printed on the grid cover,
oriented in the same direction as the grid lines and centered on the cover.
A second important properly is the grid focal distance. With most grids the tops
of the septa, closest to the x-ray source are spaced slightly closer together
than the bottoms. The idea is to orient the septa so they all point to a single
line on which the x-ray source should be located. The distance between this line
and the grid is its focal distance. X-rays coming from anywhere on this “focal’
line pass easily through all parts of the grid since the x-rays will always
intercept only the width of the septa and not their sides. X-rays coming from
any other direction are more greatly attenuated since they will hit the sides of
the septa. This is the purpose of the grid: to allow radiation to penetrate that
is coming directly from the focal spot (primary) but to preferentially attenuate
radiation coming from other directions - the scattered radiation. The side of
the grid that is supposed to face toward the focal spot is indicated by presence
of the previously discussed line that also indicated the direction of the grid
lines.
A third important properly is the grid ratio, which is the height of the lead
septa divided by their separation. The higher the grid ratio, the more selective
the grid is for radiation coming from the focal line. However the related
disadvantage of a high ratio grid is that the focal spot needs to be located
very close to the focal line or the primary radiation will be substantially
attenuated. Thus proper grid alignment becomes more critical for higher ratio
grids. Grid ratios most commonly used in radiography vary from 5:1 to 12:1.
A fourth important property is the number of grid lines per inch or per cm or
grid line frequency. Typically you will find grids with 32 to 70 lines per cm
(80 to 178 lines per inch). Grids with 32 or 40 lines per cm (80 or 102 lines
per inch) will generally produce visible grid lines when imaged with screen-film
systems, unless they are moving during the exposure. “Fine line” grids with 60
or 70 lines per cm (152 or 178 lines per inch) will generally not produce
visible grid lines and can be used without a moving grid or “bucky” mechanism.
With digital image receptors such as flat panels or CR, it is essential to use
high ratio grids if they are stationary since the lower ratio grids will produce
image artifacts called moiré patterns, due to interference or “beating” between
the grid line frequency and the digital sampling frequency. This frequently look
like diagonal stripes across the image.
All of the grid properties discussed so far are usually specified by the
manufacturer and labeled on the grid. However a fifth property usually is not.
It is the actual thickness of the lead septa, which is related to a
specification of the lead content of the grid in units of mg/cm2. Thicker septa
provide greater attenuation of radiation hitting the sides of the septa,
especially at higher kV techniques. However they also increase the absorption of
the primary radiation, which is a disadvantage. Thinner septa are generally
required for fine line grids to prevent excessively high primary attenuation.
With all of these specifications, how do you specify/select an optimal grid for
an application? Here are some basic guidelines. For grids employed in stationary
x-ray equipment in which a fixed SID (source to image receptor distance) is used
with the grid and careful grid alignment is possible, a 12:1 grid is optimal. It
provides excellent scatter rejection for increased image contrast. For grids
used in portable radiology with cassettes, use very low ratio grids. Since you
will be imaging over a range of SID’s and you cannot provide precise alignment
of the grid, you need a grid that is more forgiving - that allows the focal spot
to deviate from the grid's focal line without serious artifacts. With a 5:1 grid
the focal spot can deviate by up to about 3 inches laterally from the focal line
without serious problems. You also have a greater range of SID’s that can be
used without serious artifacts - you don’t have to be very close to the grid’s
actual focal distance. In practice grids are usually labeled with a “focal
range” over which they can be effectively used, rather than the actual grid
focal distance. Often a 5:1 grid will have a focal distance of about 52 inches,
but be labeled as having a focal range of 40 to 72 inches, which is realistic.
The focal distance labeling on a higher ratio grids however is often very
misleading. I have frequently seen 10:1 or 12:1 grids labeled with a 40 to 72
inch focal range, which is physically impossible. Such higher ratio grids can
only have a focal range of about 60 to 72 inches for grids with longer focal
distances, and about 36 to 40 inches for grids with shorter focal distances. The
rationale of having such a grid with a claimed focal range of 40 to 72 inches is
to provide convenient use for with upright cassette holder. Here one may desire
to image chests at an SID of 72 inches and do most other imaging at 40 inches.
Chests should be done at 72 inches (or greater) to reduce the image
magnification, thus allowing a greater fraction of the patient population to fit
on a single film and also to prevent different magnifications of the lungs and
heart. Spine or abdomen images however require a much higher mAs for proper
exposure and imaging these at 72 rather than 40 inches would require the mAs to
be 3.2 times higher. This would lead to longer exposure times and more patient
motion blurring, unless a rather high end imaging system with high power
capabilities were used. Yet for upright cassette holders it is actually most
common to use a grid at both 72 and 40 inches.
What happens with the high ratio "40 to 72 inch" grids is a substantial falloff
in density from the vertical centerline of the grid toward the sides, when used
at either 40 or 72 inches. Along with this falloff in density, the image
constant and primary to scatter ratio is degraded. Additionally if the grid is
at all tilted right to left or if the focal spot is not properly aligned to the
grid, the high-density region will move off to one or the other side of the
image. The result is one lung appearing darker than the other. This can easily
happen if someone bumps into the upright cassette holder or if the grid is not
set up properly initially. Some of these grids are focused midway between 40 and
72 inches so that this problem is equally severe at both 40 and 72 inches. Some
grids are actually 60 to 72 inch grids that have been relabeled and that perform
well at 72 inches but are terrible when used at 40 inches.
This definitely compromises image quality, but how do you solve this problem?
Unfortunately there are no easy solutions, but I can recommend 3 solutions.
For a wall cassette holder:
1. Select a 12:1 grid focused at 60 to 72 inches (this is the typical 72" grid).
Do chests at 72 inches and other films at 56 inches. Why 56? It just makes the
conversion of techniques from a 40 inch SID simple: multiply the mAs by 2.0
(56/40 squared). The grid will look fine at 56 inches.
2. Have 2 wall cassette holders with 2 different grids, one focused at 40" and
the other at 72".
3. Have 1 wall cassette in which the grid can be easily accessed and removed and
then switch from 40" to 72" grids depending on the SID desired. The problem here
is remember to switch the grids and not dropping/damaging them.
An obviously poor solution is to use a 6:1 grid that CAN have a focal range of
40 to 72 inches. The scatter cleanup here will be much less than a 12:1 grid and
will result in greatly decreased image quality.
The solution for a table grid is obvious: only select a 12:1 grid focused at 40
inches.
Dr Frank Ranallo, PhD, DACR is Associate Professor of Medical Physics and
Radiology at the University of Wisconsin Medical School
| |
|
