Dose Reduction Techniques in CT

First, we need to briefly discuss how one can change dose in CT. The major user-accessible parameters are mA, kV, and pitch. We have already discussed pitch. Increasing mA increases the x-ray flux through the patient with consequent increased dose and decreased noise. Increasing kV increases both the x-ray flux and the average x-ray energy with a consequent increase in dose. While noise drops with increased kV, contrast also decreases as the predominating tissue interaction is incoherent scattering rather than the photoelectric effect.

Thus the first - and the traditional - way of adjusting the dose from a CT exam is to simply adjust the mA and kV (and pitch) to the optimal values for that particular examination, with as low dose as possible for acceptable image quality. As mentioned above, the bowtie filter will always be in place and will also provide some dose reduction (and needs to be selected for the relevant body part).

However, it is often difficult to correctly estimate the parameters for a given patient, since these are strongly affected by patient body habitus. Thus, CT manufacturers developed automatic exposure compensation (AEC) methods for CT, which typically involve tube current (mA) modulation. The CT scanner will automatically choose the mA for you based on the scout (topogram) image. The mA will change dynamically along the length of the scan based on the attenuation at that z-position on the topogram.

The red line illustrates the mA determined by a typical tube current modulation technique. Notice how the mA increases around the shoulders and around the pelvis, where the bony structures cause substantial attenuation of the beam. mA decreases in the region of the lungs, which do not attenuate the beam very much.

The importance of the scout scan cannot be underestimated. If the topogram does not include that portion of the scan, the scanner does not know what dose to use and will likely use a default, high dose. Additionally, if the patient is positioned incorrectly in the bore, the topogram might be magnified, and thus the scanner will think the patient is larger than he or she is - and thus deliver a higher dose.

As the patient is positioned vertically higher in the scanner bore (right), the topogram obtained with the tube anterior is magnified. This can be seen with the similar triangles, as discussed in more detail in the section on magnification in radiography. A larger topogram will make the scanner think the patient is larger, and the AEC will deliver a higher dose. The opposite effect would be seen if the patient were lower in the bore (or if the tube were posterior, etc.).

As an additional or alternate method of dynamic mA modulation, many scanners will adjust the mA for the next tube rotation based on how attenuating the tissue was through the last rotation. Separately, most CT scanners will adjust the mA through their rotation angle - depending on how thick the patient is in that direction. In particular, most patients are thicker side-to-side than front-to-back; thus, the scanner will ramp up the mA when the tube is shooting across the patient and decrease it as the tube comes to the vertical angles. Finally, as an even more sophisticated dose-reduction technique, some scanners (e.g. Siemens with CareDose 4D) will reduce the dose across the anterior chest and increase mA through the back of the thorax. This is intended to reduce the breast dose from the scan (but will still obtain diagnostic images since the mA is increased along the equivalent, posterior rays).

In the horizontal angles, the scanner increases the mA (darker red) to penetrate the bony structures and longer effective patient diameter. As the tube becomes more vertical, there is less attenuation and so the mA can be decreased (lighter gray).

How does the AEC decide what mA to use? The fundamental concept is very simple - you want the resulting CT image to look as good in the fat patient as in the skinny patient, right? The scanner will measure the attenuation and size of the patient (at each z-position) based on the topogram and choose an mA to compensate for the size versus the 'standard patient' (and based on the user-entered 'reference' value for image quality). Interestingly, it turns out that radiologists can tolerate noisier images in large patients (perhaps because their body fat provides inherent contrast); thus, some tube current modulation technologies actually increase the mA not quite as much as you would need to give equivalent image quality.

kV and Dose Modulation. The effects of changing kV on patient dose become even more complex in the setting of automated exposure control and with large variations in patient body habitus. CT images can tolerate substantial decreases in kV in small, especially pediatric, patients (e.g. 80 kV), in whom even the lower energy x-rays can penetrate the patient. However, these low energy beams are problematic in very obese patients, in whom they simply cannot penetrate the torso. Thus, the AEC ramps up the mA to compensate. Overall, the exact dose to the patient depends highly on particular scanner settings and should be adjusted to obtain diagnostic images with as low dose as possible.