Chest: Multidetector Row Computed Tomography and 3D Volume Rendering For Adult Airway Imaging
Leo P. Lawler, MD, FRCR, Frank S. Corl, MS,
Elliot K. Fishman MD,FACR.
INTRODUCTION
The new high quality volume data sets that can be rapidly acquired with current multidetector row CT together with the latest advances in volume rendering software are making routine three-dimensional airway imaging a reality. This exhibit seeks to demonstrate the techniques used to optimally image the central and peripheral airways with illustrations of specific clinical applications.
The exhibit will cover the following areas-
A description of the technique used to harness the potential of multidetector row CT image acquisition. Various volume rendering protocols will be illustrated.
A demonstration and discussion of some of the pertinent anatomical features of normal airway anatomy.
Illustration of a series of cases that demonstrate the application of these techniques to clinical practice.
TECHNIQUE
The parameters discussed here reflect our experience with Siemens Volume Zoom multidetector row CT and 3DVirtuoso workstation (Siemens, Iselin, NJ) though the principles are generally applicable.
Most airway imaging may be performed as a non-contrast study. Contrast may be considered when an airway abnormality may be related to a vascular ring or sling. When the patient is being considered for transbronchial biopsy the location and course of mediastinal vessels may also be of value. The standard study coverage is from the thoracic inlet to below the diaphragm. On occasion it may be necessary to extend coverage to the aero-digestive tract of the neck.
The high pitch and fast gantry rotation of multidetector row CT permit rapid coverage which minimizes or eliminates the effects of breathing mis-registration. Narrow collimation may be employed without compromise of z-axis coverage. At present 1mm detector collimation is routinely employed to depict airway detail to the outer third of the lung. 2.5mm detectors can be selected for faster coverage and less noisy images. A high-resolution edge-enhancing reconstruction kernel will better demonstrate the sharp airway interfaces though on occasion may introduce excessive noise on 3D renderings. Although dynamic inspiratory and expiratory airway imaging with multidetector row studies has shown promise for tracheobronchomalacia [1]the value of such studies must be weighed against the increased dose required. As with all CT studies the radiation dose must be kept as low as possible without sacrificing diagnostic detail. The lungs and airway with their natural contrast offer potential for continued reduction in radiation dose.
Once the study is completed the images are sent to a volume rendering workstation which provides a platform for real-time image volume editing. The voxels are of high fidelity to the original data with histograms displaying all representative density components[2]. The air-soft tissue interface of the airways is ideally suited to the volume rendered technique[3, 4]. Various trapezoid algorithms may be applied to demonstrate findings to best effect with individual opacity and brightness assignments to the airway and the air it contains as well as surrounding tissues. Color assignments may be used to demonstrate airway stents and mediastinal vasculature to better effect. Slab clip-plane editing rapidly removes any overlying soft tissues. Real-time interactive change in projections offers infinite views that may be customized to the airway of interest. Editing may be used to isolate specific dichotomous branches from adjacent airways. During clinical consultation particular approaches of interest to the bronchoscopist my be simulated.
The specific clinical scanning protocol that we routinely use is;
KV/mAs/Time per rotation (s)
140/100/0.5
Collimation/slice/reconstruction (mm)
1or 2.5/1.25or 3/1
Table speed (mm per rotation) / Pitch
6/6 or 15/6
Note. Pitch is defined as table movement per gantry rotation divided by single slice collimation.
NORMAL ANATOMY
Trachea.
The intrathoracic length of the trachea is 6-9cm. It is oval at the cricoid and round below with variations in shape with respiration. The trachea scaffolding is comprised of C-shaped cartilage rings completed by the pars-membranacea posteriorly.
Bronchi.
The right bronchus is more anterior than the left. The right principle bronchus is shorter, wider and straighter than the left, which lies more horizontal. Only the right upper lobe bronchus arises outside the hilum whereas all others arise after entering the hilum. Each of the segmental bronchi supply a similarly named bronchopulmonary segment (10 total in each lung). The superior segment bronchus is the highest to arise posteriorly. The right and left segments are largely similar except that the left upper lobe has an apicoposterior segment and the left lower lobe has a small medial segment that has a common stem with the anterior basal segment.
o Right upper lobe bronchus. This bronchus is just below the carina arising more superiorly than the left upper lobe bronchus.
o Bronchus intermedius. Arising just beyond the right upper lobe bronchus this bronchus is 3cm long, posterior to right pulmonary artery and medial to right interlobar pulmonary artery. Parenchyma extends medially to it as the azygoesphageal recess. The posterior wall is contact with the superior segment right lower lobe.
o Right middle lobe bronchus. This branch courses caudally and anteriorly and is at the same level as the right lower lobe bronchus with the medial and lateral segment bronchi are located more caudally. The superior segment bronchus right lower lobe may arise at the same level as the right middle lobe bronchus.
o Left upper lobe bronchus. The left upper lobe bronchus originates at a level more caudal than the right upper lobe bronchus with the left pulmonary artery arching over it. The opening of the left upper lobe bronchus is large. The aerated superior segment of the left lower lobe abuts the posteromedial wall of the left upper lobe bronchus. The anterior segment bronchus is the only one to course directly anteriorly.
o Left apical-posterior segment bronchus. The origin of the left apical-posterior segment bronchus is at the level of the right upper lobe orifice.
o Lingular bronchus. This bronchus arises from the undersurface of the distal left upper lobe bronchus with an oblique anterior and caudal course.
o Left superior segment bronchus. The left superior segment bronchus is usually at the same level as on the right side.
o The left anterior and medial segment bronchi arise as a single structure.
Bronchovascular relationships.
o The truncus anterior/right upper lobe pulmonary artery is anterior to right upper lobe bronchus.
o The right pulmonary artery crosses anterior to the bronchus intermedius and lies more lateral as it becomes the right interlobar pulmonary artery.
o The right superior pulmonary vein lies anteromedial to the middle lobe bronchus
o The middle lobe pulmonary artery is between the medial and lateral segment bronchi
o The apicoposterior bronchus may lie medial or lateral to the branch of the left upper lobe pulmonary artery.
o The left pulmonary artery between left mainstem and apicoposterior bronchus is posterolateral to the left superior pulmonary vein
o The left superior pulmonary vein is in front of the left upper lobe bronchus, becoming more horizontal more caudally.
o The left pulmonary artery extends caudally posterior to the left interlobar pulmonary artery.
o The left interlobar pulmonary artery lies in the bifurcation between lingular bronchus and anterolateral to lower lobe bronchus.
o The right epiarterial upper lobe bronchus is superior to pulmonary artery. The remaining bronchi are hyparterial(below the bronchi). This helps in differentiating situs conditions.
o The aygous vein arches over the right mainstem bronchus.
Lung Roots.
Left lung root.
The left bronchus is below and behind the left pulmonary artery. There is a pulmonary vein anterior and inferior.
Right lung root.
The upper lobe bronchus and artery are above the main bronchus in the root. The two pulmonary veins are in front and below the main bronchus.
CLINICAL APPLICATION
Since the demise of bronchography, bronchoscopy has become the standard for airway evaluation. Though invasive, the ability to directly visualize the airway, to biopsy tissues, to lavage and place stents have made this an irreplaceable tool for the thoracic surgeon or pulmonologist. However when what is required is an evaluation of the site and degree of stenoses or their follow-up there is clearly a role for a non-invasive study such as a focused airway CT study[3-6]. Similarly when one needs to see beyond a tight stenoses or to evaluate the relationship of an extraluminal etiology there is a role for dedicated airway CT .
There is an increasing use of airway stents for benign and malignant stenotic diseases. Such stents require frequent bronchoscopic follow-up and adjustment[7]. CT of the airway may offer an easier way to monitor such cases until adjustment requires direct intervention.
In many cases the salient CT findings of the airway can be determined by close attention to the airway on the two-dimensional axial imaging review. However when information on the length and distribution of stenoses or bronchiectatic change is required the two dimensional perspective is insufficient. In-plane airways and complex branching patterns cannot be fully appreciated or assimilated by review of multiple slices and pre-procedure 'bronchoscopic maps' are hard to generate. Irregular and variable luminal diameter changes are difficult to estimate on sequential images, which may be oblique to the airway. Likewise when the concern centers on describing the relationship to an extrinsic compression of the airway a customized perspective can show its effect to best advantage.
In our experience CT evaluation of airway compromise from direct tumor invasion is the most common aetiology. Primary lung tumors, their sites of recurrence or adenopathy frequently compress or invade and distort the airway. Polyps and primary tracheal tumors are distinctly less common. The main clinical questions asked of the study are; the site and length of the stenosis, its degree and its distance from the upper trachea. The also want to know if the post obstructed lung is salvageable and whether it is likely to re-inflate after laser or stenting. It is also important to elucidate the relationship of the vasculature when bronchoscopic procedures to resect tumor are contemplated. The next most common etiology we see is airway narrowing at an anastamosis for lung transplantation related to ischemia or rejection[8]. Inflammatory conditions of the airway are unusual though we have used this technique to assess strictures related to Wegener's granulomatosis and caustic ingestion. Bronchiecatic changes are usually related to cystic fibrosis in children or traction changes of chronic interstitial fibrosis in adults.Abnormal airway dilatation can be hard to appreciate bronchoscopically. One value of the volume rendered approach is that one can better gain an appreciation of the distribution of disease from customized coronal reconstructions.
The following cases illustrate our application of multidetector row CT and volume rendering techniques to patients referred to use in routine clinical practice.
REFERENCES
1. Gilkeson, R.C., et al., Tracheobronchomalacia: dynamic airway evaluation with multidetector CT. AJR Am J Roentgenol, 2001. 176(1): p. 205-10.
2. Calhoun, P.S., et al., Three-dimensional volume rendering of spiral CT data: theory and method. Radiographics, 1999. 19(3): p. 745-64.
3. Remy-Jardin, M., et al., Volume rendering of the tracheobronchial tree: clinical evaluation of bronchographic images. Radiology, 1998. 208(3): p. 761-70.
4. Remy-Jardin, M., et al., Tracheobronchial tree: assessment with volume rendering--technical aspects. Radiology, 1998. 208(2): p. 393-8.
5. Fleiter, T., et al., Comparison of real-time virtual and fiberoptic bronchoscopy in patients with bronchial carcinoma: opportunities and limitations. AJR Am J Roentgenol, 1997. 169(6): p. 1591-5.
6. Kauczor, H.U., et al., Three-dimensional helical CT of the tracheobronchial tree: evaluation of imaging protocols and assessment of suspected stenoses with bronchoscopic correlation. AJR Am J Roentgenol, 1996. 167(2): p. 419-24.
7. Mehta, A.C. and A. Dasgupta, Airway stents. Clin Chest Med, 1999. 20(1): p. 139-51.
8. Ward, S. and N.L. Muller, Pulmonary complications following lung transplantation. Clin Radiol, 2000. 55(5): p. 332-9.