Applied Radiology

The technology of helical CT differs significantly from conventional CT,1 and its introduction into clinical practice has required a reappraisal of the basic concepts of CT scanning.2 Helical scanning is more flexible than conventional scanning but, as has been found with MRI, this flexibility results in more complex imaging protocols which must be individualized to ensure the generation of clinically appropriate data. Important technical parameters, such as collimation, table increment, and scanning time, need to be defined by the radiologist prior to the examination, and an understanding of the principles of reconstruction intervals, interpolation algorithms, and intravenous contrast dynamics is required for optimal results.3

Helical scanning has been made possible by the development of slip ring technology for gantry design. This enables continuous gantry rotation and data acquisition with simultaneous translation of the patient through the gantry at a constant rate.1 Most systems are third generation geometry scanners with a continuously rotating tube and detector system, which takes approximately one second for a 360° rotation. The x-ray focus performs a helical motion relative to the patient and from the resulting volumetric data acquisition, a contiguous set of axial images is obtained via a process of interpolation. These images are obtained, without an interscan delay, during a single breath-hold, which is in contrast with conventional CT which produces discontinuous axial images obtained during separate intervals of breath holding. The raw data can be analyzed retrospectively, at arbitrary operator-selected levels, using various interpolation algorithms.4

A number of important advantages over conventional CT result from this helical scanning technique (table l). Firstly, total scaning time is significantly reduced due to the ability to acquire multiple slices in a single breath-hold. In most patients, it is possible to scan the entire thorax in less than a minute, which is important both for patient comfort and when scanning critically ill or injured patients. Acquisition of a contiguous set of images eliminates the problem of respiratory misregistration, which has important implications in the detection of small pulmonary nodules.5,6 Furthermore, the ability to retrospectively reconstruct overlapping images improves resolution in the z-axis direction, reduces problems due to partial volume averaging, and allows more accurate characterization of small pulmonary lesions.7

Overlapping reconstructions also enhance the quality of multiplanar and 3D reformats, which have particular applications in imaging the tracheobronchial tree and the major vessels within the thorax. A further benefit of the reduction in scanning time is the more precise delivery of intravenous contrast media, which allows smaller volumes to be used and complete studies to be obtained at optimum vessel opacification.8

There also are a number of limitations to helical scanning which relate to the power capacity of the x-ray tube and image processing. Significant heat is built up during continuous acquisition, and as a result, the duration of exposure and tube current may be limited. The increased volume of data produced by helical scanning places significant demands on data storage space and image reconstruction, particularly for 3D techniques, which can be time intensive for the operator. Compared to conventional CT, there is reduced resolution in the longitudinal direction (z-axis). This results in partial volume averaging and blurring of longitudinal reformats, which is due to the broader slice sensitivity profile of helical scans.9 However, a substantial improvement has been made in this area by the use of 180° interpolation algorithms instead of 360°, and although image noise is increased in these algorithms, it does not adversely affect the diagnostic capacity of helical CT.4

Several artifacts related to helical scanning may be found. Flow artifacts due to incomplete mixing of opacified and unopacified blood are not uncommon. Often seen at the confluence of the innominate veins, their appearance may mimic intravascular thrombus. Another artifact, the stair-step, is peculiar to helical CT,10 and is typically seen in 2D and 3D reconstructions at high contrast oblique interfaces which are near the axial plane (figure 1). Stair-step artifacts occur as a result of the interpolation process and are most prominent if the reconstruction interval is much less than the collimation, which results in highly overlapping axial images.

Aortic pulsation artifact is sometimes seen in the ascending aorta and may mimic an aortic dissection. It results from a combination of z-axis blurring and aortic motion and occurs because the scan duration closely approximates the duration of a single cardiac cycle. Reducing the pitch can minimize this artifact; alternatively, segmentation of the scan with partially reconstructed images can help clarify the situation.11

2D and 3D reformatting

Reformatting techniques add to the value of helical CT, and technological advances in software design are making the scans easier to produce. There are a number of methods of manipulating the volumetric data set, and the ones most commonly employed in thoracic imaging are multiplanar reformats, shaded-surface displays (SSD), and maximum or minimum intensity projections.12 Each has its own advantages and limitations.

Multiplanar reformats can be performed relatively easily from highly overlapping reconstructed images, and the operator may select any number of sagittal, coronal, or oblique planes. Curved planes are often best for displaying pathology, although they result in distortion of anatomy.

SSDs and intensity projections require image editing to remove unwanted densities, such as bone and soft tissue, before postprocessing can be performed. SSDs are based on a density threshold technique that only displays objects within the preselected density range. This display gives an excellent depiction of anatomy and overlapping structures, due to the perception of depth, in a format which surgeons find particularly helpful as an aid to surgical planning. It is, however, time intensive and operator dependent, and the threshold technique often results in loss of information, in particular internal details like plaque calcification and intimal flaps. The degree of a stenosis also may be overestimated with this technique.

Intensity projections are generated by mapping the maximum or minimum attenuation values obtained from passing multiple imaginary rays through the image data set. Unlike SSDs, intensity projections preserve relative densities, although overlapping structures are less well displayed. This limitation can be overcome by generating images in different degrees of orientation about a single axis. Minimum intensity projections have a unique application in the thorax due to the low density of air in the lungs and airways, and they provide good depiction of the tracheobronchial tree, although the technique again tends to overestimate stenoses.

Indications and techniques

We routinely use helical CT scanning for all patients referred for chest CT. However, there is no uniformly agreed upon method for a standard helical CT study: 13our current protocol for a thoracic survey uses a collimation of 7 mm with a pitch of 1.5 (table 2). Scans are obtained from the thoracic inlet to the kidneys using two or three 15-second breath-holds; we find the majority of patients can manage this easily. Intravenous contrast is not needed in all cases but, when used, is administered as a bolus of 100 ml of 60% iodine at 2 ml/sec, with scanning commencing after a 20-second delay. Reconstruction uses a standard algorithm for the mediastinum and lung parenchyma. The reconstruction interval is equal to the collimation, with overlapping reconstructions reserved for situations where patient management may be significantly altered, i.e., excluding further lesions in a patient with a solitary nodule. Patients who are being evaluated for interstitial lung disease have a thoracic survey scan without contrast, using a 10 mm slice thickness with 10 mm spacing and a pitch of 1.5, followed by 1 mm slices at 10 mm intervals with targeted reconstruction in a high frequency algorithm (table 3).

In addition to improved detection and characterization of pulmonary nodules,7,14 the key areas where helical CT has impacted in thoracic imaging are in the assessment of the tracheobronchial tree and examination of the thoracic vasculature, in particular the pulmonary arteries and aorta. Detailed analysis of the tracheobronchial tree is required for tumors, strictures, patients with hemoptysis and, more recently, lung transplant patients to assess the anastamosis for stenosis or dehiscence.15

Our technique uses a 3 mm slice thickness with a pitch of 1.5 and reconstruction at 3 mm intervals. The axial images supplemented by coronal reformats can assess the degree and length of a stenosis, aid significantly in surgical planning, and provide a road map

for transbronchial biopsy and endobronchial procedures such as stenting (figure 2).16 Visualization beyond stenoses is an added advantage in situations where bronchoscopy has been unable to cross the narrowing.

Improved depiction of vascular structures within the chest is possible with helical CT due to its combination of fast scanning and more uniformly consistent opacification of vessels.2,8 Particular interest centers on the assessment of aortic aneurysms and dissections,18 and the diagnosis of pulmonary thromboembolic disease;19 there also is interest in its potential role in the evaluation of coronary artery bypass graft patency.20 No standard method for optimal contrast enhancement has been established. Some radiologists use a test injection to determine peak enhancement of the vessel of interest and thus calculate scan delay for the main study, while others intuitively use a scan delay of around 20 to 30 seconds, depending on the injection rate. The latter method is appropriate for slower injection rates (2 to 3 ml/sec), while the test injection technique is more useful at higher rates (3 to 5 ml/sec). At our institution, we use 150 ml of contrast, injected at 3 ml/sec after a 20 second delay (table 4).

Both 2D and 3D reformats provide useful information in assessing the extent of aortic aneurysms (figure 3) and dissections (figure 4). Helical CT has been shown to be equal to both MRI and transesophageal echo in the diagnosis of aortic dissection, and it has been shown to reduce the need for angiography in assessing patients with aortic trauma (figure 4).

A diagnosis of acute pulmonary thromboembolism requires optimal opacification of the pulmonary vessels. In our experience, 100 ml of contrast injected at 3 ml/sec after a 15 second delay provides excellent images when used with 3 mm slices at 2 mm intervals (figure 5) (table 5). Remy-Jardin et al19 found that helical CT reliably demonstrated emboli in 2nd to 4th division pulmonary arteries with an overall sensitivity of 100% and specificity of 96%, compared to pulmonary angiography. However, false positives may occur due to intersegmental lymph nodes, and small peripheral emboli may not be detected. The role of helical CT in the assessment of chronic pulmonary emboli is less well defined. Although helical CT may not yet entirely replace angiography in the diagnosis of acute thromboembolism, it has an important role in the follow up of patients with documented central emboli, and is a powerful alternative for patients who are at high risk for the complications of pulmonary angiography.

Conclusion

Helical CT offers significant advantages over conventional CT in the imaging of pulmonary nodules, the tracheobronchial tree, pulmonary arteries, and the aorta. Advances in helical techniques now enable imaging of vascular structures, the results of which have been found to be equal to that of conventional angiography. At present, conventional pulmonary arteriography and aortography may be abdicated in many cases because helical CT, with adequate bolus contrast administration, can provide equal diagnostic capability in a noninvasive manner. With further advances in helical techniques and image processing, helical CT will most likely replace conventional angiography for many diagnostic vascular examinations. AR

References

1. Kalender WA, Seissler W, Klotz E, et al: Spiral volumetric CT with single breath-hold technique, continuous transport and continuous scanner rotation. Radiology 176:181-183,1990.

2. Heiken JP, Brink JA, Vannier MW: Spiral (helical) CT. Radiology 189:647-656,1993.

3. Brink JA: Technical aspects of helical (spiral) CT. Radiol Clin North Am 33(5):825-841,1995.

4. Polacin A, Kalender WA, Marshal G: Evaluation of section sensitivity profiles and image noise in spiral CT. Radiology 185:29-35,1992.

5. Collie DA, Wright AR, Williams JR, et al: Comparison of spiral acquisition computed tomography and conventional computed tomography in the assessment of pulmonary metastatic disease. Br J Radiol 67:436-444,1994.

6. Krudy AG, Doppman JL, Herdt JR: Failure to detect a 1.5 cm lung nodule by chest computed tomography. J Comput Assist Tomogr 6:1178- 1180,1982.

7. Costello P, Anderson W, Blume D: Pulmonary nodule: Evaluation with spiral volumetric CT. Radiology 179:875-876,1991.

8. Costello P, Dupuy DE, Ecker CP, et al: Spiral

CT of the thorax with reduced volume of contrast material: A comparative study. Radiology 183:663-666,1992.

9. Brink JA, Heiken JP, Balfe DM, et al: Spiral CT: Decreased spatial resolution in vivo due to broadening of the section-sensitivity profile. Radiology 185:469-474,1992.

10. Wang G, Vannier MW: Stair-step artifacts in three-dimensional helical CT: An experimental study. Radiology 191:79-83,1994.

11. Posniak HV, Olson MC, Demos TC: Aortic motion artifact simulating dissection on CT scans. Elimination with reconstructed segmented images. AJR 161:557-558,1993.

12. Fisherman EK, Magid D, Ney DR, et al: Three-dimensional imaging. Radiology 181:321-337,1991.

13. Paranjpe DV, Bergin CJ: Spiral CT of the lungs: Optimal technique and resolution compared with conventional CT. AJR 162:561-567,1994.

14. Remy-Jardin M, Remy J, Giraud F, et al: Pulmonary nodules: Detection with thick section spiral CT versus conventional CT. Radiology 187:513-520,1993.

15. Schaeffer CM, Prokop M, Ziak C, et al: Spiral CT of anastamotic complications after lung transplantation [abstr]. Radiology 189:263,1993.

16. Mergo PJ, Fenton J, Helmberger T, et al: Presurgical and preprocedural applications of 3D CT in the tracheobronchial tree [abstr]. AJR 164 (American Roentgen Ray Society 95th Annual Meeting Program Book Suppl):240,1995.

17. Kauczor H-U, Wolke B, Fischer B, et al: Three-dimensional helical CT of the tracheobronchial tree: Evaluation of imaging protocols and assessment of suspected stenoses with bronchoscopic correlation. AJR 167:419-424,1996.

18. Costello P, Ecker CP, Tello R, et al: Assessment of the thoracic aorta by spiral CT. AJR 158:1127-1130,1992.

19. Remy-Jardin M, Remy J, Wattine L, et al: Central pulmonary thromboembolism: Diagnosis with spiral volumetric CT with the single breath-hold technique: Comparison with pulmonary angiography. Radiology 185:381-388,1992.

20. Tello R, Costello P, Ecker C, et al: Spiral CT evaluation of coronary artery bypass graft patency. J Comput Assist Tomogr 17:253-259,1993.

Dr. Paley, Dr. Taylor and Dr. Mergo are in the Department of Radiology at the University of Florida College of Medicine in Gainesville, FL.