Nick Reading
Department of Radiology
Whipps Cross Hospital
September 1999
This article has three pages (including this one)
- Follow the links at the bottom of each page.
Page 2: Anatomy of thoracic CT...
Page 3: Putting anatomy into practice...

Basic principles and anatomy of thoracic CT

When performing a CT scan of the chest there are several choices to be made:

(1) Standard, high resolution or CT angiography scan
(2) Whether to give contrast
(3) Windowing
(4) Reconstructions

(1) Standard, high resolution or CT angiography

These types of scan demonstrate specific things and are not interchangeable.

Standard scan covers the entire chest in a series of fairly thick slices varying from 5-10mm depending of the type of scanner available. It therefore gives a complete overview of the chest and is used for staging lung cancer, looking for pulmonary metastases and assessing the mediastinum. If the scan is performed on the older style "sequential" scanner which acquires each slice individually, then small portions of the lung can be missed if the patient does not suspend breathing in the same phase for each slice. The more modern spiral (or more correctly, helical) CT scanners overcome this problem by scanning the entire chest in one breath-hold.

High resolution scan is specific for diffuse lung disease. Thin slices (1-2mm) are taken every 10-20mm, processed with a bone algorithm (which basically adds edge enhancement) and viewed on lung windows (see below). This technique gives superior detail of the lung parenchyma in terms of small airways, interlobular septa etc. and allows classification of diffuse lung disease. It is important to realise that only a small percentage of the chest is imaged by this technique and so it would be quite useless for staging lung cancer. In addition, the bone algorithm severely degrades detail in the mediastinum. Complete coverage of the lungs with these thin slices is not possible due to the consequent high radiation dose.

Comparison of 10mm slice (a.) and 1mm high resolution slice (b.) in patient with coal worker's pneumoconiosis

CT angiography is only possible with a spiral CT scanner due to its speed of image acquisition. A large volume (100-150ml) of contrast is injected into a peripheral vein at a steady rate (c.3ml per sec) and the scan timed to image the main pulmonary vessels while the contrast bolus is in the pulmonary arteries. Fairly thin slices are obtained. Detailed analysis of the images allows accurate diagnosis of many cases of pulmonary embolism.

CT pulmonary angiogram. Thrombus in pulmonary arteries, outlined by contrast, is indicated by arrows

Thicker slices allow greater coverage of the great vessels and provides a quick and accurate method of assessing possible aortic injury or dissection.

CT angiogram. Axial image, (a.) and sagittal reformat, (b.), show dissection flaps in both ascending and descending aorta.

(2) Contrast

Intravenous contrast in thoracic CT is used mainly to demonstrate vessels. This is not always necessary as the anatomy of many vessels in the chest is so constant that they are obvious on the unenhanced scan. However, problems may be encountered in the mediastinum and hilar regions in distinguishing lymph nodes and other soft tissue masses from vessels and in these cases intravenous contrast can be invaluable. Contrast medium should be administered with care in asthmatic patients for fear of provoking an acute asthma attack. Use of contrast is contra-indicated in patients allergic to iodine. Contrast can precipitate acute renal failure in patients with impaired renal function and such cases should be discussed with the radiologist in advance.

(3) Windowing

Windowing is something that is done after the scan has been performed and so can be changed at a later date if necessary, provided one has access to the CT scanner. Windowing is like very sophisticated contrast and brightness controls. The amount of information in the CT image is too great to be presented completely on the monitor or film. Because of this, a range of contrast and brightness is selected by the radiographer for optimal demonstration of the structures one is interested in. In order to see lung parenchyma optimally completely different windowing will be needed compared to that used for the mediastinum. It follows that one must have the correct windowing to assess specific tissues.

 

Contrast enhanced CT scan of chest displayed on soft tissue windows. Fat, muscle and bone are easily distinguished by their different levels of grey. Lung detail is not displayed with these window settings.

 

CT scan at approximately the same level displayed on lung windows. Lung vessels, bronchi and fissures are well seen but muscle and fat cannot be differentiated. Bone is hard to distinguish from soft tissue.

(4) Reconstructions

Because spiral CT acquires a volume of data rather than individual slices multiplanar reconstructions and 3D volume techniques can be used to produce images that look very impressive. Generally the most accurate data is obtained by viewing the original axial slices, but occasionally certain reconstructions can be of use. There is some evidence that 3D reconstructions of the pulmonary vessels provides a good method for screening for pulmonary AVM's. Special computer programmes also allow one to perform virtual bronchoscopy, but it is worth remembering that the images presented to the virtual endoscopist are produced from the original axial images so that anything present in the endoscopic tour is generally seen more accurately on the original axial slices. You don't get something for nothing!

3D surface shaded display of pulmonary AVM, (a.) before treatment, (b.), after coil embolisation.

Small arrow: feeding artery, Large arrow: draining vein, Open arrow: embolisation coil

Page 2:
Anatomy of thoracic CT...

 

heppell.net