Imaging the Lung: Is Pattern Recognition Overemphasized?
World Small Animal Veterinary Association World Congress Proceedings, 2004
Donald E. Thrall, DVM, PhD, DACVR (Radiology, Radiation Oncology)
College of Veterinary Medicine, North Carolina State University
Raleigh, NC, USA


Students of imaging consider the lung to be one of the more difficult regions to interpret proficiently. This relates to the wide range of normal lung appearance. Normal lung appearance is affected by radiographic technique, ventilation characteristics of the patient, and by differences in breed and body conditioning.

Technical quality cannot be compromised when making thoracic radiographs. Low mAs and high kVp combinations should be used to overcome the high inherent patient contrast created by the large volume of air in the lung. In some practices, sedation is routinely used when patients are radiographed and the resultant decrease in ventilation will result in an overall increased opacity to the lung that will invariably be over-interpreted as interstitial disease.

In veterinary imaging it has been standard practice to categorize pulmonary disease into compartments, i.e., bronchial, interstitial and alveolar... the so-called "pattern recognition" method. In actuality, involvement of just one of these compartments in any pulmonary disease is unlikely and this commonly used classification then becomes problematic. Students struggle with categorizing pulmonary opacification and perhaps the "pattern recognition" method is not optimal for all patients. A slight variation in this classic approach could prove useful; it involves first emphasizing whether the radiograph is actually abnormal and then making a decision whether airways or alveoli are actively involved with the disease process or whether the primary disease is mainly confined to the interstitium. On the surface this sounds identical to the "pattern recognition" system, but it varies in that students from the outset are concerned with normal vs. abnormal (the most difficult of all assessments) and then whether airways are actively involved. Strict classification into anatomic compartments is avoided.

Many normal thoracic radiographs are misinterpreted as abnormal. The main reason for this is poor radiographic technique with underexposure or poor ventilation, both of which will be misinterpreted as interstitial disease. Before the image is scrutinized for abnormalities, the following assessments should be made: 1) is the patient obese? 2) were the images acquired on expiration or has ventilation been compromised by sedation?, and 3) was the radiographic technique adequate to penetrate the thorax? When and only when the answer to all 3 questions is 'no' should the image be scrutinized for abnormalities.

A helpful tip is to compare the appearance of the lungs in the lateral views to the VD (or DV) view. Routinely there will be an overall increased opacity in the lung seen in lateral views that is interpreted as interstitial disease. This is caused by the greater degree of recumbent atelectasis that is present in lateral views. In these patients, lungs are better aerated when the patient is in sternal or dorsal recumbency and as a result they appear more radiolucent, and normal, in VD or DV views.

Infiltration of the air space with blood fluid or cells results in "alveolar" disease. Emphasis has been placed on presence of air bronchograms and the lobar sign as radiographic indicators of alveolar disease. Other characteristics such as indiscrete margins and the tendency to coalesce have also been discussed. Little attention has been paid to the intensity of the opacity on a per unit area, or the discreetness of the margins of the abnormality.

Students overemphasize the need for air bronchograms to be present for a radiographic diagnosis of alveolar disease to be made. It is a fact that air bronchograms are often not present in instances of significant alveolar disease, especially in cats; thus, other signs must be used. I promote first considering the intensity of any abnormal opacity on a per unit area. The two situations creating intense lung opacification on a per unit area are alveolar infiltration and a mass. Masses and alveolar infiltration are usually readily distinguishable based on the margins of a mass being discrete and the margins of an infiltrate being indiscrete. Visualization of air-bronchograms or the lobar sign will be further evidence for an alveolar component of the disease. However, first concentrating on the intensity of the abnormality will remove the pressure associated with attempting to identify an air bronchogram. In reality, many air bronchograms identified by students are not air bronchograms at all but simply normal lung interposed between paired pulmonary arteries and veins. Once an abnormality is considered to be "alveolar", signalment and history can be used as usual to formulate the list of rule-outs and one can consider use of a transtracheal aspirate or bronchoalveolar lavage for diagnostic purposes. The most common causes of a pulmonary infiltrate that is primarily alveolar are exudates, edema and hemorrhage.

Bronchial disease typically results from thickening of bronchial walls and/or a peribronchial infiltrate. Taken alone, these radiographic signs are not difficult to identify. However, rarely does bronchial disease exist as a sole entity and end-on views of abnormal bronchi (sometimes called ring shadows or donuts) or side-on views of abnormal airways (sometimes called tram-lines) are often seen in combination with other pulmonary opacities. Students struggle whether to call this a mixed pattern, a bronchointerstitial pattern, a bronchoalveolar pattern or just a variant of bronchial disease. In actuality it does not matter. The bottom line is that there is an airway (bronchial) component of the disease and performing a transtracheal aspirate or bronchoalveolar lavage to retrieve material that will be helpful in making the diagnosis is a reasonable consideration.

Pulmonary disease that is primarily of a bronchial nature is most often of allergic origin. The combination of bronchial disease and involvement of another component of the lung has more widespread cause including allergy, pulmonary edema, mycoses and atypical metastasis.

The interstitium is the supporting framework of the lung. Primary interstitial disease results when there is a proliferation of tissue or accumulation of fluid in this framework. Interstitial fluid presence is a phase in the development or resolution of pulmonary edema. Interstitial pulmonary edema as the only manifestation of the pulmonary edema syndrome is short lived and rarely recognized radiographically. Tissue proliferation in the interstitium is typically due to cancer and nodules forming as a result of metastatic or primary lung cancer do not usually presents a diagnostic challenge.

Other types of tissue proliferation in the interstitium due to inflammatory disease such as mycotic pneumonitis (low grade or it will also quickly involve the alveoli) or other types of cancer, such as lymphoma or metastasis from a glandular primary, will result in a diffuse unstructured or reticular increased opacity in the lung. Here again consideration of the overall intensity of the opacity on a per unit area is important as these primary interstitial diseases will not have the intensity of opacification as an alveolar infiltrate or an interstitial mass. Notably, neither air bronchograms nor lobar signs will be seen. Thus, in this instance use of a transtracheal aspirate or bronchoalveolar lavage to obtain material for diagnostic purposes may not be productive as there is either no involvement or nonspecific secondary involvement of the airways and alveoli. This latter point is important. Students often ponder thoracic radiographs wondering if there is mild generalized increased interstitial opacification. The bottom line here is that unless the patient's clinical signs are severe enough that a lung aspirate or biopsy is being considered, a final decision on whether there are mild interstitial changes is irrelevant. Most of the time the patient can be safely monitored with subsequent thoracic radiographs to characterize any change in the possible lung abnormality.

In summary, a simplified radiographic approach to interpreting the lung where effort is first directed at deciding whether the lung is abnormal followed by a primary assessment of whether the airways are involved may decrease confusion and may also be useful for deciding when other diagnostic techniques such as transtracheal aspiration or bronchoalveolar lavage should be employed.

Speaker Information
(click the speaker's name to view other papers and abstracts submitted by this speaker)

Donald E. Thrall, DVM, PhD, DACVR (Radiology, Radiation Oncology)
College of Veterinary Medicine, North Carolina State University
Raleigh, NC

MAIN : Diagnostic Imaging : Imaging The Lung
Powered By VIN