CT Radiographic Techniques
Tube Current (mA)
The x-ray tube current determines the rate at which x-rays are produced in the x-ray tube (i.e., photons per second). The total number of x-ray photons that are acquired in the 1000 or so projections as the x-ray tube rotates 360º degrees is proportional to the product of the mA and the rotation time (seconds), or the mAs. Since the x-ray tube rotation is normally fixed, the number of photons used to make any CT image is directly proportional to the tube current (mA).
|Figure A. Cross sectional images of an anthropomorphic phantom. All images were acquired and reconstructed using identical techniques except for the variations in x-ray tube current which increases from 10 mA (upper left) to 640 mA (bottom right).
Figure A shows an example of a series of images acquired at tube currents ranging from 10 mA to 640 mA, which is an increase of a factor of 64. Figure B shows a "blow up" from a nominally uniform region from selected images, and which clearly demonstrate how increasing the mA (i.e., photons used to acquire the image) reduces the amount of noise (mottle) in the resultant CT images. CT imaging is generally a quantum noise limited imaging system, where the only significant source of noise is quantum mottle, and other sources such as electronic noise are deemed to be negligible. For a quantum noise limited imaging system, the noise is inversely proportional to the square root of the number of photons used to make the image. In other words, if the number of photons used is quadrupled, the noise in the resultant image should be halved.
|Figure B. Small regions of the taken from the same phantom location, but acquired using different mA values.
Figure C illustrates quantitatively the reduction in image noise with increased mA, and illustrates that quadrupling of the selected mA approximately halves the measured noise value. Figure C also shows that the Hounsfield Unit value is independent of the mA value. Since the HU value is the amount of material contrast relative to water, this example also illustrates that changing mA does NOT affect contrast, but only noise. Image quality, and the ability to detect a lesion, is generally related to the contrast to noise ratio. Increasing the mA will improve image quality by reducing noise, but will not affect contrast.
|Figure C. Analysis of a small region of interest showing that the mean attenuation (i.e., Hounsfield Unit) is independent of the mA used to acquire the image, but that the noise (standard deviation or Std. Dev.) in the region of interest is reduced as the mA increases.
X-Ray Tube Voltage (kV)
The x-ray tube voltage is an important parameter for all x-ray based imaging modalities, including CT. When the x-ray tube voltage increases, the number of x-rays produced increase. The increase is greater than the linear relationship that is observed when the mA is changed; doubling of the x-ray tube voltage would likely increase the total number of x-ray photons produced by about a factor of four i.e., x-ray tube output is taken to be approximately proportional to kV2.
Increasing the x-ray tube voltage will also increase the average photon energy. As the photon energy increases, x-ray attenuation is reduced. The attenuation coefficients in a CT image are expressed in Hounsfield Unit (HU), which quantifies the amount of attenuation of any specified tissue relative to the attenuation of water. In general, changes in x-ray photon energy will also result in a change in the tissue HU value. If a tissue (lesion) were surrounded by water, the tissue HU quantifies the difference in x-ray attenuation of the lesion relative to the surrounding water (i.e., contrast). Changes in kV will therefore also result affect the amount of contrast in the resultant image.
Figure D shows four images acquired using four x-ray tube voltages ranging from 80 to 140 kV, but otherwise having identical techniques and acquisition geometry. The four x-ray tube voltages depicted in Figure D (i.e., 80, 100, 120 and 140 kV) are those that are most frequently encountered in clinical CT imaging. In each image depicted in Figure D, there is a region of interest (~20 cm2) that depicts the average HU value in a constant tissue equivalent material, as well as the corresponding standard deviation of the measured HU values around the mean. Increasing the x-ray tube voltage from 80 to 140 kV changes both the mean value, as well as the corresponding standard deviation:
Mean HU value: The mean HU value at 80 kV is 115 HU, and this increases to 141 HU at 140 kV. The increase in HU at 140 kV relative to that of 80 kV (i.e., 26 HU) means that relative to water, this tissue equivalent material is 2.6% more attenuating at the higher x-ray tube voltage [1 HU corresponds to a change of x-ray attenuation, relative to that of water, of 0.1%]. The behavior of the HU of any tissue depends on three factors: (a) photon energy, where increasing the photon energy generally increases the likelihood of Compton scatter interactions, and reduces the likelihood of photoelectric interactions; (b) tissue density, which is the principal determinant of x-ray attenuation at high photon energies of tissue like materials (i.e., those with atomic number similar to water [Z = 7.5]; (c) atomic number (Z), atomic numbers, where the higher the atomic number, the more likely the photon interactions are likely to be photoelectric. The variation of tissue HU with photon energy may be complex; for example, a low Z material may be less attenuating than water at low photon energies where photoelectric processes are more important, but more attenuating at high energies if the material density is high, since at high energies Compton processes are more important and these are approximately proportional to the physical density. In general, tissue like materials (i.e., Z ~ 7.5) will show only modest variation in HU with photon energy in CT, of the order of a couple of percent or so, as depicted in Figure D. High Z materials, however, like Iodine (Z = 53) and Barium (Z = 56) will show a rapid decrease in HU with increasing photon energy (kV) because x-ray interactions for these materials are dominated by the photoelectric effect, and the PD effect is inversely proportional to E3 (i.e., doubling the photon energy will reduce the photoelectric effect by about a factor of eight).
Standard deviation in HU: The data in Figure D show that increasing the x-ray tube voltage reduces the image noise/mottle, as reflected by the observed reduction in the measured standard deviation. There are two reasons for the observed reduction in noise/mottle with increasing kV: (a) as the kV is increased, (many) more photons are produced in the x-ray tube, and the number measured by the x-ray detectors will therefore be increased; (b) increasing the kV also increases the average photon energy, which increases the x-ray penetration, so that the percentage of the x-ray intensity incident on the patient that is transmitted also increases. For both of these reasons (i.e., more photons, and increased penetration), increasing the x-ray tube voltage will reduce the amount of mottle observed in the resultant images as shown by the data in Figure D. It is interesting to note that increasing the x-ray tube voltage from 80 to 140 kV reduced the noise from 8.9 HU to 4.3 HU; to achieve this twofold reduction in noise would have required a quadrupling of the mA, or scan time, at a constant x-ray tube voltage.