Irene C. Van Der Schaaf, MD

The introduction of 40-slice CT scanners has opened up new possibilities for CT angiography of the supra-aortal vasculature. Imaging can be performed with even thinner slices, and more rapidly, than on 16-slice systems, and images have higher resolution. Conventional protocols for imaging the brain and its arterial supply must be adjusted to profit from these parameters.

Shorter scan times are expected to reduce motion artifacts, provide better contrast enhancement, reduce venous contamination, and require smaller dosages of contrast material. Extended coverage is especially useful for brain perfusion, but it can capture the entire extra- and intracranial pathway of the carotid arteries, instead of just the bifurcation. High spatial resolution along the patient axis (z-axis) makes it possible to evaluate the smallest aneurysms (2 mm). This capability puts 40-slice CTA in direct competition with MR angiography.

Most indications in multislice CT involve a three-part protocol: unenhanced CT scanning, CT perfusion imaging, and then CTA. Unenhanced CT excludes intracerebral hemorrhage, extensive infarct, or hemorrhagic transformation of an infarct. If no abnormality is seen, perfusion CT may show early signs of ischemia. CTA is used as a final step to demonstrate most of the underlying vascular pathology (see table). The following details are universally applicable to many different 40-slice CT scanners:

- Unenhanced CT. This can be performed using spiral scan techniques or the axial scan mode. If an axial mode is used, the patient's ocular lenses can be kept outside the scan field by tilting the gantry. This helps avoid radiation-induced cataracts. Use of thin sections for scanning (40 x 0.625 mm) followed by reconstruction of thicker slices (5 mm) results in good signal-to-noise ratio. Thin sections (0.9 mm) can be reconstructed retrospectively if necessary. These thin axial sections can be used to reconstruct 5-mm coronal and/or sagittal sections, which may provide additional information about the skull base or regions close to the vertex.

Spiral scanning is an option for uncooperative patients because it takes only a few seconds and motion can be minimized. Another advantage is the immediate availability of thin sections. These can be used to optimize symmetric positioning of axial slices and provide additional 3D information on the location and extent of bleeding or infarcts. The patient's lenses are always included in the scan field, however, even if the console indicates they are outside the scan range, due to overranging. The CT scanner requires additional data for interpolation at the two ends of the scan range and performs approximately one additional rotation, making it impossible to exclude the lenses from the exposed range.

- Perfusion CT. Perfusion imaging is performed immediately after the unenhanced scan, which provides a baseline for the calculation of perfusion values. Time to peak enhancement of the perfusion can then serve as a test bolus for CTA. Contrast-to-noise ratio is optimized by using the lowest possible kVp values. This is of major importance when calculating perfusion maps.

 It is important to avoid the lenses, and so exact determination of scan angulation is required on the scout view. Scanning can also be performed in so-called jog mode. This feature doubles scanned volume, thereby covering almost the entire brain (4-cm coverage times two).

Dynamic perfusion images can be acquired every two or three seconds instead of every second.

 While this does not influence perfusion parameters significantly, it can reduce patients' radiation dose substantially. Thirty images are usually sufficient to cover the whole bolus for quantitative perfusion calculations. It is essential to minimize head movement; otherwise, calculation of perfusion values becomes impossible. The reconstructed slice thickness should be restricted to 5 mm. Thicker slabs suffer from partial volume effects, thereby influencing perfusion values.

A delay of about five minutes between perfusion CT and CTA is recommended. This is not essential, but it will reduce venous overprojection from contrast used in the perfusion scan.

- CTA. Imaging the circle of Willis conventionally requires 120 kVp in combination with 160 to 200 mAs. Use of lower values (80 to 90 kVp), however, results in a better contrast-to-noise ratio at equal or lower total dose.

 Arterial enhancement is brighter and depicts vascular structures better, especially in the presence of subarachnoid hemorrhage, despite increased noise. Continued use of 120 kVp is advised for carotid imaging, because the shoulder region causes too much noise at lower kVp. Introduction of dose modulation and adaptive filters may be helpful in such cases.

A caudocranial scan is generally used when imaging the circle of Willis, although imaging of the carotid arteries is performed best from the craniocaudal direction, which produces less venous overprojection and reduces scatter at the level of the subclavian artery. Optimum z-axis resolution requires the thinnest possible collimation that still covers the desired scan range. A pitch of 1 or just below will be sufficient. Lower pitches (0.3 to 0.6) can be used in cases of clipped cerebral aneurysms to reduce clip artifact. Increased kVp (140 kVp) in combination with higher concentrations of contrast (370 mg I/mL) also helps to reduce clip artifact.

Use of contrast is critical to obtaining a high degree of vascular opacification throughout the imaging volume and to displaying the vessel of interest with as few artifacts as possible. Achieving a plateau phase of enhancement is unnecessary when using 40-slice CT, given the rapid scan times.

Case-by-case determination of scan delay is mandatory with these short scan times. Perfusion CT should be used as a test bolus if venous enhancement is to be minimized, but if venous structures are of interest, then an additional wait of four to eight seconds should be considered. A saline flush can also help reduce contrast artifacts in injection veins, while increasing the potential enhancement that is required for both CTA and perfusion CT.

Image review of the carotid arteries is based primarily on curved planar reformations in the coronal and sagittal directions. Vessel tracking software helps track the vessel course automatically, but it is less useful for the petrous segment of carotid arteries. Curved planar reformations should follow vessels from the aortic arch to the intracranial portion of the internal carotid artery, then to one of the branches of the external carotid artery and to the vertebral arteries on either side. With three vessels on each side and two projections, therefore, 12 images must be reviewed. Using these 12 images makes evaluation easy and time-efficient.

Subarachnoid hemorrhage is most often (85%) caused by a ruptured intracranial aneurysm. Delay in diagnosing this life-threatening situation can be avoided by performing CTA immediately after unenhanced CT identifies the hemorrhage.

 CTA can assess aneurysmal characteristics and treatment possibilities.

 The thin slices can be evaluated by scrolling interactively through the scan volume in either multidirectional, multiplanar reconstruction, or maximum intensity projection mode. Rotating thicker MIP or volume-rendered slabs around a target lesion can display the aneurysm neck and its relationship to surrounding vessels (Figure 1).

Patients will still be at risk from complications, even after aneurysmal occlusion. Most complications are due to delayed cerebral ischemia, which can be caused by vasospasm. CTA may be used to assess vasospasm, while additional perfusion CT can provide important information regarding brain tissue at risk from ischemia.

Perimesencephalic nonaneurysmal hemorrhage, in which subarachnoid blood is centered around the midbrain, accounts for two-thirds of nonaneurysmal subarachnoid hemorrhages. CTA can help exclude aneurysm in such cases. The prognosis for perimesencephalic nonaneurysmal hemorrhage is good. Clinical and radiological characteristics suggest a venous origin. Patients often have primitive venous drainage directly into dural sinuses instead of via the vein of Galen.

 The side of the bleeding relates to the side of primitive drainage.

Patients with sinus thrombosis present with a wide variety of symptoms. This condition may occur spontaneously or as a complication of another preexisting disease. Unenhanced CT is often the first imaging examination requested. It may reveal a dense vessel sign or hemorrhagic infarct due to venous congestion, but it often shows no abnormalities. CT venography can be performed easily in the same session. Perfusion CT imaging can replace the usual bolus timing and may reveal early-stage at-risk brain tissue (Figures 2A and 2B). No waiting period between this perfusion scan and CT venography is necessary because venous enhancement is desirable. A delay of four to eight seconds should be used instead.

The best way to detect sinus thrombosis is to review the scan volume in three planes in MPR mode. Additional curved planar reformations can show the thrombus throughout the sagittal sinus. These reconstructions should be centered precisely to avoid creating thrombus- or lumenlike artifacts (Figure 2C).

CT venograms and MR venography are equally accurate for detecting sinus thrombosis. Advantages of CT in general include its widespread availability, reduced likelihood of motion artifacts, and better possibilities for patient monitoring. MR provides better visualization of the brain parenchyma, however.

Acute ischemia, or stroke, is the third most common cause of mortality in Western countries, and it also accounts for a high level of morbidity. While conservative measures have traditionally been the only options for treating stroke patients, thrombolytic therapy is now successful in 15% to 20% of cases. This approach carries a risk of complications, however, and adequate imaging is essential.

 Patients with acute stroke undergo unenhanced CT, perfusion CT, and CTA in succession. Use of the 40-slice scanner allows more brain tissue to be covered during perfusion imaging. The CTA examination will be quicker and will thus require less contrast for the carotids.

No further scans will be required if unenhanced CT shows extensive infarction or intraparenchymal bleeding. Perfusion imaging may reveal early ischemia and potentially salvageable tissue (penumbra) if few or no abnormalities are seen. Wider coverage (jog mode) can help identify the infarct location. Measures of cerebral blood volume (CBV), cerebral blood flow (CBF), and mean transit time (MTT) are generally used to define brain tissue status. Reperfusion is unlikely to save this brain tissue when absolute CBV is less than 2 mL/100 g or CBF is less than 20 mL/100 g/min. Thrombolytic therapy is considered more beneficial when a large penumbra (ischemic but viable tissue) is present (Figure 3).

The high resolution and shorter scan times associated with 40-slice CT make it possible to visualize intracranial and extracranial vessels within 10 seconds. This may eventually result in more rapid instigation of preventive treatment. Although contrast is used for both CTA and perfusion CT, the total amount required can be kept within reasonable limits.

Carotid artery stenosis is a common complication of atherosclerosis. Ultrasound is preferred for nonacute carotid stenosis screening in patients who have suffered stroke or transient ischemic attack. Detection of stenosis greater than 50% should be followed by further imaging to determine the exact degree of stenosis and the presence of tandem stenoses and/or calcifications and to assess completeness of the circle of Willis. This can all be demonstrated on 16-slice CTA as well as 40-slice CTA.

Brain perfusion in patients with carotid artery stenosis may reveal severe perfusion deficits in some patients, while others show complete symmetric perfusion, despite unilateral carotid disease. Thus, in some cases, brain perfusion may be helpful to decide whether a patient will benefit from treatment.

Stroke risk increases in patients with ulcerated plaques that facilitate formation of thromboemboli. Imaging of carotid plaque is therefore important.

 Calcified plaques complicate stenosis evaluation unless transverse sections or curved planar reformations are generated. A wide window setting should be used.

Studies investigating whether fatty or calcified plaques influence the outcome of carotid artery stent placement are ongoing. CTA can demonstrate new plaques outside the stent, as well as intima proliferation and in-stent restenosis on follow-up (Figure 4).

We have found that stenosis quantification is most precise if curved planar reformations are generated parallel to-and transverse sections perpendicular to-the artery's longitudinal axis. Curved planar reformations are especially important in the carotids' bony segment, where most other visualization techniques fail.

Precise positioning of the centerline within the vessel will prevent simulation of stenoses from eccentric cuts. Experienced practitioners can grade stenosis severity by visual estimation. Less experienced staff may choose to use calipers. We use a wide window setting in clinical patients, contrary to results from phantom experiments.

Internal carotid artery dissections occur in young and middle-aged patients alike. They are a major cause of stroke (5% to 20%) in the young population. Nontraumatic dissections may occur spontaneously or after trivial trauma such as rapid head turning or a normal sporting activity. Stroke is often delayed for days or weeks while the dissection remains extracranial. Anticoagulation therapy is usually effective, and 85% of patients with minor symptoms do well. Only 5% of patients suffer severe strokes.

The most common extracranial dissection site is a few centimeters beyond the bifurcation in the internal carotid artery. Carotid dissection may present as a smooth, long stenotic segment that extends intracranially. This is known as the string sign. The false channel is usually thrombosed and may appear hyperdense with a crescent shape. A thin rim of enhancing vasa vasorum may be seen around the intramural hematoma (Figure 5). Diagnosis is easy with CTA if the false channel is perfused, but high-resolution imaging is essential for optimal display of such a channel. Dose modulation is obligatory given the high proportion of young patients.

Traumatic dissection is more likely to involve the common carotid artery and result in complete occlusion. Patients with traumatic dissection present more frequently with cerebral ischemia and respond less well to therapy. Comparison with the contralateral side is crucial for diagnosis. Carotid CTA in trauma patients can be combined with chest and abdominal scanning, if required.

Forty-slice CTA has become a competitive technique for evaluation of the intra- and extracranial arteries. The high spatial resolution, larger coverage, and reduced requirement for contrast material make the combination of CTA with perfusion CT an excellent strategy for many acute and subacute situations. Important issues that remain and should be considered are options for dose reduction and optimization of image review.

DR. WAAIJER and DR. VAN DER SCHAAF are radiology residents, DR. VELTHUIS is a radiologist, and PROF. PROKOP is a professor of radiology, all at University Medical Center Utrecht in the Netherlands.

1. Wintermark M, Uske A, Chalaron M, et al. Multislice computerized tomography angiography in the evaluation of intracranial aneurysms: a comparison with intraarterial digital subtraction angiography. J Neurosurg 2003;98(4):828-836.

2. Velthuis BK, van Leeuwen MS, Witkamp TD, et al. Surgical anatomy of the cerebral arteries in patients with subarachnoid hemorrhage: comparison of computerized tomography angiography and digital subtraction angiography. J Neurosurg 2001;95(2):206-212.

3. Wintermark M, Maeder P, Verdun FR, et al. Using 80 kVp versus 120 kVp in perfusion CT measurement of regional cerebral blood flow. AJNR 2000;21(10)1881-1884.

4. Wintermark M, Smith WS, Ko NU, et al. Dynamic perfusion CT: optimizing the temporal resolution and contrast volume for calculation of perfusion CT parameters in stroke patients. AJNR 2004;25(5):720-729.

5. Bahner ML, Bengel A, Brix G, et al. Improved vascular opacification in cerebral computed tomography angiography with 80 kVp. Invest Radiol 2005;40(4):229-234.

6. Cademartiri F, van der Lugt A, Luccichenti G, et al. Parameters affecting bolus geometry in CTA: a review. JCAT 2002;26(4):598-607.

7. Fleischmann D, Hittmair K. Mathematical analysis of arterial enhancement and optimization of bolus geometry for CT angiography using the discrete fourier transform. JCAT 1999; 23:474-484.

8. Roos EJ, Rinkel GJ, Velthuis BK, Algra A. The relation between aneurysm size and outcome in patients with subarachnoid hemorrhage. Neurology 2000;54(12):2334-2336.

9. Velthuis BK, van Leeuwen MS, Witkamp TD, et al. Computerized tomography angiography in patients with subarachnoid hemorrhage: from aneurysm detection to treatment without conventional angiography. J Neurosurg 1999;91(5): 761-767.

10. van der Schaaf IC, Velthuis BK, Gouw A, Rinkel GJ. Venous drainage in perimesencephalic hemorrhage Stroke 2004; 35(7):1614-1618.

11. Wardlaw JM. Overview of Cochrane thrombolysis meta-analysis. Neurology 2001;57(5 Suppl 2):S69-S76.

12. Smith WS, Roberts HC, Chuang NA, et al. Safety and feasibility of a CT protocol for acute stroke: combined CT, CT angiography, and CT perfusion imaging in 53 consecutive patients. AJNR 2003;24(4):688-690.

13. Klotz E, Konig M. Perfusion measurements of the brain: using dynamic CT for the quantitative assessment of cerebral ischemia in acute stroke. Europ J Radiol 1999;30(3):170-184.

14. Schramm P, Schellinger PD, Klotz E, et al. Comparison of perfusion computed tomography and computed tomography angiography source images with perfusion-weighted imaging and diffusion-weighted imaging in patients with acute stroke of less than 6 hours' duration. Stroke 2004;35(7):1652-1658.

Irene C. Van Der Schaaf, MD

Articles

40-slice scanners boost neuro CT angiography