Applied Radiology

Approximately 10 years after spiral CT revolutionized CT imaging, a technical breakthrough occurred that has proven to be as, or more, important. Multislice CT (MSCT) offers better longitudinal and temporal resolution that benefits nearly all imaging applications. The first MSCT systems were introduced in the early 1990s and were composed of dual- or split-detector systems permitting acquisition of two channels of X-ray data simultaneously. MSCT scanners with four data channels were introduced in 1998 and have proven to be a significant advancement in CT performance, providing narrower slices, faster scans, and/or greater longitudinal coverage. In 2001 to 2002, systems with 8 or 16 data channels have emerged and manufacturers are testing even broader detector arrays as a future means of providing "snapshot" CT imaging. Dissemination of this technology has been rapid and widespread. However, the potential for increased radiation exposure with this technology has dampened enthusiasm, particularly in the pediatric population.

The width of the radiation profile with MSCT is increased substantially, due in part to the shape of the beam. MSCT systems use a cone beam, rather than a fan beam, geometry. Although the radiation beam width is typically within 1 millimeter of the nominal scan width for single-detector CT scanners, the width of the radiation beam typically exceeds the total scan width in MSCT scanners. The amount of beam width overage that occurs is dependent upon the detector configuration used. In the earliest version of a 4-channel MSCT scanner (Lightspeed QX/i, Version 1.0, GE Medical Systems, Milwaukee, WI), the radiation profile width was 150% greater than the scan width (12.5 mm) for a 5-mm nominal scan width (4 × 1.25 mm detector configuration). Conversely, for a 20mm nominal scan width (4 × 5 mm detector configuration), the radiation profile width was only 30% greater than the scan width (26 mm). ¹ As a result, the maximum surface CT dose index (CTDI) values for multislice body CT increased by 238% for the 4 × 1.25 mm detector configuration and 76% for the 4 × 5 mm detector configuration, as compared with single-detector CT. Thus, the dose efficiency of MSCT is maximized when the full longitudinal extent of the detector is used (4 × 5 mm detector configuration).

The difference between single-slice and MSCT for CT examination of the adult abdomen and pelvis were well documented by McCollough and Zink. ¹ Examining the differences between their single-detector protocol (7-mm slice thickness, pitch of 1) and their MSCT protocol (5-mm slice thickness, beam pitch of 0.75), these authors found that the scan time was reduced from 34 to 16 seconds for 30 cm of coverage with a rotation time of 0.8 seconds. Moreover, although image noise was equivalent between the two techniques, the tube current was reduced from 310 mA to 190 mA with MSCT, resulting in a total mAs of 10,540 for single-detector CT as compared with 3,040 for MSCT. However, two factors counterbalanced this reduction in tube current such that the radiation dose increased by approximately 50% at both the center and the surface of a 32-cm CTDI phantom. Specifically, the increased width of the radiation profile, and the 25% overlap in the X-ray beam with spiral MSCT (beam pitch = 0.75) both contributed to this apparent paradox. In spite of these differences, the authors point out that the absolute dose from the MSCT examination is well within the acceptable values for comparable single-detector CT examinations.

In a later version (Lightspeed QX/i, Version 1.1, GE Medical Systems), a technique was developed to reduce such dose inefficiencies. A focal-spot tracking algorithm was designed to compensate for unwanted motion in the X-ray tube that results from both thermal and mechanical effects. By creating a feedback loop from the detector to cams at the X-ray beam collimator, portions of the X-ray beam that fell beyond the active detector cells owing to beam motion were greatly reduced. As a result, the maximum surface CTDI values increased by just 105% for the 4 × 1.25 mm detector configuration and only 10% for the 4 × 5 mm detector configuration as compared with single-slice CT. Although the maximum surface dose is increased by 10% with MSCT, the effective dose (a measure related to the total energy deposited in the patient) is nearly equalized between the two techniques. This is because the overlap between scans performed during separate breath-holds with single-slice helical CT is eliminated with MSCT performed in a single breathhold.

With single-slice helical CT, a well-known dose benefit may be realized by increasing the beam pitch. Radiation dose may be cut in half by doubling pitch from one-to-two pitch = 2.0, without significant loss of diagnostic information. Similarly, with MSCT, one may consider using a beam pitch of 1.5 for routine imaging as the 25% overlap with beam pitch of 0.75 is replaced with a 50% gap in the X-ray beam. However, the system software of most MSCT scanners automatically adjusts the tube current in order to obtain comparable levels of image noise, which largely offsets any potential benefit to radiation dose, unless one manually overrides this adjustment and accepts an increased level of image noise. ²

One of the primary benefits of MSCT is the ability to image with thinner slices. But, as thinner sections are acquired, radiation dose is necessarily increased to maintain photon flux, so long as image noise is held constant. Although this is true for both single-slice CT and MSCT, the use of narrow-beam collimation with MSCT imparts a relative dose inefficiency owing to the increased percentage of the X-ray beam that falls beyond the active detector rows (penumbra). To compensate for this, one must lower the tube current and accept an increased amount of image noise among the thin sections, a practice that is usually acceptable, since thin sections are typically used for high-contrast imaging applications such as CT angiography. Review of thicker sections for routine image viewing may be performed by generating thicker sections from a thin-slice acquisition, or by postprocessing thin slices into thicker reformations on an image review workstation. Either technique will compensate, in part, for the increased noise associated with acquisition of thinner slice data.

Manufacturers are developing new innovations for radiation dose reduction. Beyond focal-spot tracking techniques intended to minimize wasted radiation from the penumbra, manufacturers continue to develop techniques that permit the radiation dose associated with any given CT scan to be tailored to the patient's unique body habitus. Taking advantage of differences in patient thickness between the anteroposterior direction and the mediolateral direction, some manufacturers have elected to modulate the X-ray beam intensity as the beam rotates around the patient. With such an approach, substantial dose reduction benefits may be realized. Similarly, the beam may be modulated according to longitudinal differences in patient thickness as the patient travels through the X-ray gantry. When used together, transverse and longitudinal beam modulation techniques work synergistically to reduce the radiation dose.

With the widespread dissemination of MSCT scanners, manufacturers have sought to extend the technology to higher numbers of data channels that may be active at any given time, increasing the number of slices that may be acquired simultaneously. MSCT scanners that made use of a matrix detector configuration where detector elements are all of equal size have had a slight advantage in extending their scanner design to permit acquisition of 8 slices owing to the configuration. Manufacturers that used an adaptive array generally required redesign of the detector array to permit acquisition of a larger number of slices. However, matrix detectors tend to be less dose efficient than adaptive arrays, owing to attenuation of the X-ray beam by the numerous septae that divide the individual cells in the detector array. Because of this and other geometric considerations, most 16-channel implementations have converged to use of an adaptive detector array.

With an increase in the number of data slices acquired simultaneously from 4 to 8 or 16, manufacturers realize a dose efficiency benefit owing to a decrease in the amount of wasted radiation that results from the penumbra, which extends beyond the active detector rows. This is realized because of the greater longitudinal coverage achieved with each rotation of the X-ray tube permitting fewer instances in which this geometric inefficiency occurs. Siemens Medical Systems (Iselin, NJ) reports an increase in dose efficiency from 70% for 4 × 1 MSCT to more than 85% with 16 × 1.5 MSCT.

Practically, several practice and protocol adjustments can greatly offset the potential increase in radiation dose associated with spiral MSCT. ³ First, one should take care to review all current protocols to be sure that diagnostic needs are met without excess radiation. ⁷ Second, the slice thickness should be appropriate for the clinical question. Thinner sections are accompanied by increased noise, and there is a tendency to compensate for this with increased radiation. In addition, the beam width (and detector configuration) should be chosen to be as wide as possible because dose efficiency decreases as beam collimation narrows. Although increasing pitch will reduce dose with sin-gle-slice CT, the same is not necessarily true for MSCT. Some MSCT manufacturers automatically increase tube current to produce equal noise when pitch is increased. ² Radiation dose will only be lowered if the technologist manually overrides this feature. Of course, a similar dose reduction could be achieved by holding pitch constant and reducing mAs. (Some manufacturers have opted to normalize the tube cur-rent with respect to pitch and refer to the normalized value as the "effective" tube current.) Finally, technique charts should be used in the pediatric population to insure that the chosen tube current is appropriate for the patient's age and body habitus. ^4-6