Intravascular contrast material administration is essential in CTA (6). Both technical factors and patient-related factors affect the process of contrast enhancement. It is important to achieve homogeneous intravascular enhancement during imaging. Arterial enhancement depends on iodine concentration of the contrast medium used and the injection protocol (9). More enhancement is achieved with higher iodine concentration. Faster injection rates rather than increased contrast medium volume also provide more enhancement, because a higher concentration of iodine is injected over time (9).

Several contrast injection protocols have been used for CTA. Uniphasic injection protocols with fixed scan delay, injection volume, and injection rate usually result in sufficient aortic enhancement (10). An interactive protocol, in which
the injection of contrast medium is stopped manually, can be used to reduce the volume of contrast medium administered (11). Uniform, plateau-like arterial enhancement can only be achieved using a biphasic injection protocol (12). A biphasic injection protocol starts with a small bolus with a high flow rate, followed by a larger bolus with a lower flow rate (12). A saline flush is recommended to flush the veins and push the contrast column into the circulation to reduce the amount of contrast agent (6). Contrast injection for abdominal imaging is preferably administered in a cubital vein.

Major patient-related factors determining contrast enhancement are body habitus and cardiac status. In case of an abdominal aortic aneurysm (AAA), enhancement can be inhomogeneous because of turbulent flow (Fig. 34.1). Because of these and other differences in circulation dynamics between patients, scanning with a fixed delay is not recommended (12). In addition contrast volume and injection rate need to be adjusted according to patient weight to achieve optimal vascular enhancement. Solutions to optimize scan delay for determination of the optimal enhancement phase in a single patient are bolus triggering and the use of a test bolus. With the first technique, a region of interest is placed in the distal descending aorta for abdominal CTA. During contrast injection, low-dose nonincremental scans are obtained in a dynamic fashion. The attenuation is monitored in the region of interest and when a predefined enhancement threshold is reached, for example, 150 HU, the CT scan starts automatically with a delay of a few seconds, depending on the type of CT scanner (6). The second technique uses a test bolus to determine patients’ contrast transit time. A region of interest is placed in the distal descending aorta for abdominal CTA and during injection of 15- to 20-mL contrast medium, the test bolus, a series of low-dose nonincremental CT scans are obtained. An enhancement curve is then automatically calculated, and the time to maximum enhancement equals the mean contrast transit time. The time to maximum enhancement is set as the scan delay. However, scanning should be started after an extra delay of approximately 10 seconds, because with a large contrast bolus maximum enhancement is achieved later compared with a test bolus because of the accumulation of the contrast agent in the blood vessels (6).

With shorter acquisition times, the amount of contrast agent to be injected can be reduced, but higher injection rates are required for optimal enhancement. When parenchyma imaging is also required (e.g., in liver imaging), the volume of contrast medium cannot be reduced, because enhancement of the parenchyma depends on the total iodine dose (9).

Oral contrast administration is not advisable for vascular imaging; superposition of bowel loops filled with contrast medium can hamper the evaluation of the contrast-enhanced blood vessels, especially when reconstructions are obtained. However, when oral contrast administration is required for imaging, negative contrast medium (e.g., water) can be used, which will not disturb evaluation of the contrast-enhanced blood vessels. In addition, the use of negative oral contrast agents facilitates assessment of the bowel wall when bowel ischaemia is suspected.

Vascular imaging of the abdominal aorta and its side branches requires a high-resolution protocol to produce a (near) isotropic dataset of thin overlapping cross-sectional images. Two-dimensional (2D) and 3D reconstructions of high quality can be produced from this dataset. With 4- or 16-row MDCT, section thickness of 1 to 1.25 mm and 1 mm, respectively, can be obtained for high-resolution scanning. With thinner section thickness, scanning is not fast enough for covering the entire body volume of interest, whereas with 64-row MDCT, section thickness of 0.5 mm can be obtained routinely.

Our standard scan protocol for the Aquilion 64 (Toshiba Medical Systems, Toshiba cooperation, Tokyo, Japan) for the abdominal aorta and mesenteric vessels is performed in the supine position, with the arms elevated. Scanning is performed during a breath-hold. Section thickness is 0.5 mm, helical pitch is 53, and reconstruction index and interval are 1 and 0.8 mm, respectively. The selected kV is 120, the milliampere varies per patient because tube current modulation is used, and the rotation time is 500 milliseconds. For an average-weighted patient, arterial imaging is performed with an injection of 100-mL contrast agent (370 mg iodine/mL) at 4 mL/s, followed by 40 mL of saline flush at the same injection rate with at least 18G venous access, injected with a power injector (Stellant injector, Medrad, Medrad Inc., Indianola, PA). Contrast agent volume and flow is adjusted according to patient weight. Delay is determined by bolus triggering, and scanning starts with a delay of 10 seconds when an enhancement increase of 100 HU is reached.

For the renal arteries slice thickness is 0.5 mm, helical pitch is 53, and reconstruction index and interval are 1 and 0.8 mm, respectively. Tube current modulation is used resulting in different milliampere per patient; kV (120), and rotation time (500 milliseconds) is unchanged. Parameters can be changed individually depending on the patient’s weight.

Standard vascular protocols for the Aquilion 320 (Toshiba Medical Systems) routinely use a slice thickness of 0.5 mm. Helical pitch is 65, and reconstruction index and interval are 1 and 0.8 mm, respectively.

A great challenge of MDCT is in dealing with a large data load (13), especially if an isotropic or near isotropic acquisition is performed (3). For example, routine multislice CTA of the abdominal aorta with a scan range of 30 cm performed with 1-mm collimation and a reconstruction index of 1 mm and reconstruction interval of 0.8 mm produces 375 images. Using 64-slice CT scanners, collimations of 0.5 mm can be achieved and thus a reconstruction index of 0.5 and a reconstruction interval of 0.4 can be used. Using these parameters, 750 images are generated for the same scan volume. However, in our opinion, using slice thickness of 0.5 does not substantially add to the diagnosis in abdominal CTA and we therefore use a reconstruction index and interval of 1 and 0.8 mm as indicated earlier in 64- and 320-slice CT. Imaging the peripheral arteries generates even more images, because the scan volume is much larger. Workstation-based review is necessary not only to evaluate the axial images, which still remains the primary mode for evaluating the abdominal aorta (14), but also to generate reconstructed and 3D images (13). Several reconstruction techniques can be used, which allows better evaluation of the entire vasculature. The main visualization techniques are multiplanar reformation (MPR), maximum intensity projection (MIP), and volume rendering (VR) (15).

An MPR is a 2D reconstruction in any coronal, sagittal, or oblique plane. MPR presentation is easy to use, but restricted to a single 2D plane. Therefore it is of limited value for vascular imaging, especially in tortuous arteries (Fig. 34.2). A more sophisticated technique is curved planar reformation (CPR). With CPR reformation, the display planes curve along an anatomic structure through the entire dataset (16), for example, the abdominal aorta and a single iliac artery (Fig. 34.3), projected in a 2D image. This technique is a valuable tool for vascular imaging. We use a Vitrea workstation (Vital Images Incorporated, Plymouth, MN), which
can generate CPRs semiautomatically. At the same time, additional perpendicular images are displayed, allowing better evaluation of diameters, stenosis, and thrombus composition, thereby improving visualization of eccentric lesions.

MIP presentation visualizes only the brightest structures in a selected plane (e.g., anteroposterior) and provides an angio-like image presentation (Fig. 34.4) (14). Thick- or thinslab MIP images can be obtained. Thin-slab images allow better visualization of complicated anatomic structures. A limitation of the MIP technique, because it is a projection technique, is that more attenuated structures like bones obscure the contrast-enhanced blood vessels (13). Editing to remove overlying structures from the MIP images is usually time consuming. Another limitation of MIP presentation is the absence of the appreciation of depth relationships, especially in regions with a complex anatomy (13). MIP presentations are sometimes inadequate for visualizing small vascular structures, for example, accessory renal arteries in living-related kidney donors. Variants of MIP are minimum intensity projections and curved-slab MIP. The latter can include multiple vessels in a single image with improvement of the interpretation efficiency (17).

VR is the most complex technique (Fig. 34.5). With VR, every voxel value is assigned an opacity level, ranging from opacity to transparency. The opacification function can be applied to a selected region in the histogram of voxel opacity values, thereby visualizing the selected tissue of interest, for example, blood vessels (13). Colors can be applied to the attenuation histogram, to differentiate voxel values and, for instance, to distinguish a calcified plaque from enhancing arterial lumen. VR presentations can be used for virtual endoscopy to evaluate the internal surface of tubular structures (13,18). Mostly basic VR presentations are provided by dedicated post-processing software without prior editing. In addition, smaller regions of interest can be isolated from the scan volume manually by using dedicated software.

Additional visualization techniques like MIP and VR presentations are most often used in vascular imaging. The interpretation of an examination of the abdominal aorta takes approximately 20 minutes, using a dedicated workstation. Evaluating only 3D images is not an option because of the limitations that have been discussed. It is always important to review the cross-sectional images to obtain additional information (6).