求助英译汉，急！

1 X-ray Beam Filtration
The use of an absorbent material between the x-ray tube and the patient can be used to “harden”
the beam such that low-energy x-rays (which contribute disproportionately to absorbed dose) are
reduced, or to “shape” the x-ray beam to deliver dose in the most appropriate spatial distribution.
Previously, only head and body beam shaping (e.g., “bowtie”) filters were available. Recently,
manufacturers have added filters more specific to cardiac imaging or different-sized patients.
2 X-ray Beam Collimation
The use of a very attenuating material between the x-ray tube and the patient should be used to limit
the x-ray beam to the minimal dimensions required. Such collimation occurs along the z-axis to
define the radiation beam width. Additional collimation after the patient to further define the image
width, whether an absorbent material or electronic, causes radiation dose to the patient to be wasted.
Finally, the fan angle of the beam should be collimated to the diameter of the patient to reduce the
amount of bypass that can then be scattered back towards the patient or towards personnel. Such inplane
beam collimation is typically implemented by use of an appropriate scan FOV (shaping filter).
MDCT systems have been observed to have a radiation dose inefficiency at narrow beam collimations,
resulting in a higher CTDI for the narrow beam collimations required for narrow slice
widths. In SDCT, CTDI is generally independent of slice width(although for some SDCT systems,
the CTDI can increase by as much as a factor of 2 for scan widths less than 2 mm).
The dose inefficiency in current MDCT designs is due to unused x-ray beam that strikes outside
of the active area of the detector (along the z-direction). The z extent of this unused portion
of the x-ray beam is approximately constant in size for the various detector configurations; thus
the inefficiency caused by the unused radiation is relatively greater at narrow beam collimations.
In 4-channel MDCT systems, the narrow beam dose inefficiency can be substantial, resulting in
as much as a 40% to 50% dose increase for the narrow beam collimations (4x1 mm or 4x1.25
mm) relative to the widest beam collimations (4x5 mm or 4x8 mm). For submillimeter beam collimations
on 4-channel MDCT systems, this dose increase can be over 100% relative to the widest
beam collimations. The use of a greater number of data channels (16 or more) covering larger zaxis
extents of the detector increases the dose efficiency of MDCT to nearly that of SDCT.
3 X-ray Tube Current (mAs) Modulation and Automatic Exposure Control (AEC)
It is technologically feasible for CT systems to adjust the x-ray tube current (mA) in real-time during
gantry rotation in response to variations in x-ray intensity at the detector, much as fluoroscopic
x-ray systems adjust exposure automatically. This capability, in various implementations, is available
commercially on MDCT systems in response to wide interest from the radiology community. Some
systems adapt the tube current based on changes in attenuation along the z-axis, others adapt to
changes in attenuation as the x-ray tube travels around the patient. The ideal is to combine both
approaches with an algorithm that “chooses” the correct tube current to achieve a predetermined
level of image noise.
By decreasing or increasing the x-ray tube current, the radiation output of the tube is proportionately
changed. Image noise is dominated by the noisiest projection (which corresponds to the most
attenuating paths through the patient). Hence data acquired through body parts having less attenuation
can be acquired with substantially less radiation without negatively affecting the final image
noise. This principle can be applied to modulate the mA angularly about the patient (anteriorposterior
[AP] vs. lateral) as well as along the z-axis (neck vs. shoulders); the tube current can also
be modulated within the cardiac cycle (systole vs. diastole), or with respect to sensitive organs (PA
vs. AP)。
With regard to cardiac CT, the radiation dose for a retrospectively gated exam, where the x-ray
tube is kept continuously on throughout the acquisition, can be dramatically decreased if the tube
current is reduced during portions of the cardiac cycle that are not likely to be of interest for the
reconstructed images. Thus, in addition to modulation of the tube current based on patient attenuation,
the tube current can be modulated by the ECG signal. Since cardiac motion is least during diastole
and greatest during systole, the projection data are least likely to be corrupted by motion artifact
for diastolic-phase reconstructions. Accordingly, the tube current is reduced during systole. Dose
reductions of approximately 50% have been reported using such a strategy. The implementation of
these and other dose reduction strategies is expected industrywide over the next several years, in
response to the strong concern about the radiation dose from CT from both the public at large and
the medical community.
In addition to technical methods of dose reduction, investigators are working to determine clinically
acceptable levels of image noise for a variety of diagnostic tasks. That is, high-contrast exams(e.g., lung, skeletal, colon, sinus) require much less dose (can tolerate higher noise levels) compared
to low-contrast exams (e.g., brain, liver, and other abdominal organs). If the required noise level can
be predefined, CT systems can use technical approaches to deliver the minimum dose required to
achieve the specified noise level. The definition of a robust and standardized noise metric is required,
however, to allow a manufacturer-independent method of prescribing the desired image quality.
4 Size-or Weight-based Technique Charts
Unlike traditional radiographic imaging, a CT image never looks “overexposed” in the sense of
being too dark or too light; the normalized nature of CT data (i.e., CT numbers represent a fixed
amount of attenuation relative to water) ensures that the image always appears properly exposed. As
a consequence, CT users are not technically compelled to decrease the tube-current-time product
(mAs) for small patients, which may result in excess radiation dose for these patients. It is, however,
a fundamental responsibility of the CT operator to take patient size into account when selecting the
parameters that affect radiation dose, the most basic of which is the mAs.
As with radiographic and fluoroscopic imaging, the operator should be provided with appropriate
guidelines for mAs selection as a function of patient size. These are often referred to as technique
charts. While the tube current, exposure time, and tube potential can all be altered to give the appropriate
exposure to the patient, in CT users most commonly standardize the tube potential ( kVp) and
gantry rotation time (s) for a given clinical application. The fastest rotation time should typically be
used to minimize motion burring and artifact, and the lowest kVp consistent with the patient size
should be selected to maximize image contrast.
Although scan parameters can be adapted to patient size to reduce radiation dose, it is important
to remember certain caveats when contemplating such adjustments. First, body regions such as the
head do not vary much in size in the normal population, so modification of scan parameters may not
be applicable here based on head size.
Numerous investigators have shown that the manner in which mA should be adjusted as a function
of patient size should be related to the overall attenuation, or thickness, of the anatomy of interest
as opposed to patient weight, which is correlated to patient girth, but not a perfect surrogate as a
function of anatomic region. The exception is for imaging of the head, where attenuation is relatively
well defined by age, since the primary attenuation comes from the skull and the process of
bone formation in the skull is age dependent.
Clinical evaluations of mA-adjusted images have demonstrated that radiologists do not find the
same noise level acceptable in small patients as in larger patients. Because of the absence of adipose
tissue between organs and tissue planes, and the smaller anatomic dimensions, radiologists
tend to demand lower noise images in children and small adults relative to larger patients. For
body CT imaging, typically a reduction in mA (or mAs) of a factor of 4 to 5 from adult techniques is
acceptable in infants58. For obese patients, an increase of a factor of 2 is appropriate. For head CT
imaging, the mAs reduction from an adult to a newborn of approximately a factor of 2 to 2.5 is appropriate.
Sample technique charts are provided in appendix A. To achieve increased exposure for obese
patients, either the rotation time, or the tube potential, may also need to be increased.
5 Detector Geometric Efficiency
Ideally, all of the photons that pass through the patient should be used in the image formation
process. However, the conversion of photon energy to electrical signal is not a 100% efficient process
(although it exceeds 90% for modern scintillating detectors). New detector materials having even
MEASUREMENT, REPORTING, AND MANAGEMENT OF RADIATION DOSE IN CT

higher absorption and conversion efficiencies are of course desirable, as are detector and signal processing
electronics with very low inherent noise levels. Additionally, the small detector elements
are divided along the detector arc and along the z axis with radiation-absorbing septa (walls).
These septa also provide essential optical isolation between detector elements, but they, along with
the very fine signal transmission wires, create “dead spaces” in the detector and hence waste radiation
dose. As detectors continue to be divided into smaller and smaller discrete elements, the geometric
efficiency of the detector systems must be maintained. One important step in reducing the dead
space has been to attach and route the signal transmission wires underneath each detector element,
instead of between detector elements. However, ongoing reductions in voxel size will likely be limited
by the exponential increase in image noise that would accompany such changes61.
6 Noise Reduction Algorithms
Data processing can be performed on the raw data (in sinogram space) or on already reconstructed
images to reduce image noise. A variety of approaches are possible, all of which seek to smooth out
random pixel variations (noise) while preserving fine detail and structure (signal). With a successful
noise reduction scheme, an image of adequate quality can be acquired with a reduced patient
dose