Principles of Radiographic Interpretation of the Thorax

Donald E. Thrall

The acquisition of thoracic radiographs is common in small animal practice but is less common in the horse. Computed tomography (CT) is very useful for characterizing intrathoracic disease in dogs and cats but is not used for thoracic imaging in horses. Magnetic resonance imaging is used rarely in veterinary medicine for assessing the thorax.

Obtaining good-quality radiographs of the thorax is relatively easy, even in the horse, as long as a few basic principles are followed. However, poor attention to detail will result quickly in degradation of image quality. Given the complexity of the thorax and the multiple structures that are visible, any degradation in image quality will lead to a disproportionate loss of diagnostic quality.

For many radiographic examinations, two orthogonal projections provide adequate information for a thorough radiographic assessment. However, for the thorax there are special circumstances that make the routine two-orthogonal-projection approach either suboptimal (dogs and cats) or impractical (horse).

Nomenclature

The naming of lateral thoracic radiographs made with the patient recumbent and a vertically directed x-ray beam, as commonly done in dogs and cats, violates the point-of-entrance to point-of-exit nomenclature system described in Chapter 5. In that system, a lateral view made with the dog or cat in left recumbency would be termed a right-left view because the x-ray beam strikes the right aspect of the thorax and exits the left aspect of the thorax. But, for the thorax, the terminology for lateral projections has been abbreviated to describe the side of the patient lying on the x-ray table. Thus, for example, a thoracic radiograph acquired with the dog or cat lying on the left side would be termed a left lateral view.

In the horse, where lateral thoracic radiographs are usually made with the patient standing and a horizontally directed x-ray beam, the point-to-entrance to point-of-exit system should be used. Thus, lateral radiographs should be described as either left-right or right-left views.

In all species, ventrodorsal (VD) and dorsoventral (DV) radiographs of the thorax are named according to the point-to-entrance to point-of-exit system.

Positioning: Dog and Cat

The basic thoracic radiographic examination in the dog and cat should consist of left lateral, right lateral, and either VD or DV projections. The need for obtaining left and right lateral views routinely is explained later. A case can be also be made for acquiring VD and DV thoracic radiographs routinely, in addition to the left and right lateral views, and a rationale supporting this is also provided later.

The effort required to complete a thoracic radiographic examination is obviously related to the number of projections. There is no reason to limit the number of projections acquired routinely, especially with a direct digital system where the image is available instantly and there are no expendable supplies. Failure to acquire both right and left lateral views routinely in addition to a VD or DV view will result in some diagnoses being missed. Admittedly, the increased number of exposures per patient has the potential to increase the occupational radiation dose to personnel, but as long as proper protection principles are followed, this potential increase will not be realized.

It is important that the entire thoracic cavity be included in the image. The field of view should extend from just cranial to the thoracic inlet, to a few centimeters caudal to the last rib. The dorsocaudal tips of the caudal lobes extend further caudally than often anticipated, and these regions can be omitted easily from the field-of-view, increasing the chance for missing disease in this region. Even in the largest dogs, including the entire thoracic cavity in one 14-inch by 17-inch image is usually possible. For the occasional huge dog where this is not possible, then each projection (left lateral, right lateral, VD, DV) should be divided into cranial and caudal portions. This effectively doubles the requisite number of images to evaluate the entire thoracic cavity in huge dogs.

For lateral and VD/DV views, the thoracic limbs should be pulled cranially to remove the brachial muscles from being superimposed on the cranial aspect of the thorax. Failure to do this will cause an artificial increase in radiographic opacity of the cranial aspect of the thorax and lead to errors in interpretation of the cranial mediastinum and cranial lung lobes.

Lateral Views

Atelectasis

When an animal is in lateral recumbency, as for a lateral radiographic examination, the dependent lung collapses from (1) the effects of the heart compressing the lung, (2) reduced movement of the dependent thoracic cage because of compression, and (3) cranial excursion of the dependent portion of the diaphragm. This collapse occurs in awake, sedated, and anesthetized dogs and cats relatively shortly after being placed in lateral recumbency. The magnitude of this collapse, which is related directly to the body mass of the patient, can be appreciated by making a radiograph of the thorax with the dog in lateral recumbency and using a horizontally directed x-ray beam. Such a projection would be termed a right (or left) recumbent horizontal-beam VD. The configuration for acquiring such a radiograph and an example are provided (Fig. 25-1). The amount of recumbency-associated atelectasis that occurs is striking. This atelectasis cannot be visualized directly in a conventional lateral radiograph made with a vertically directed x-ray beam because the aerated nondependent lung is superimposed on the collapsed dependent lung. Nevertheless, the increased opacity of the dependent lung contributes to an appearance of overall increased lung opacity in lateral radiographs that is misdiagnosed commonly as disease.

Fig. 25-1 A, Schematic illustration of the geometry used to acquire a horizontal-beam ventrodorsal radiograph of the thorax. For this figure, a transverse CT image of the thorax was used to represent the dog being radiographed. D, dorsal; V, ventral. B, Horizontal-beam ventrodorsal thoracic radiograph acquired using the geometry depicted in A. Note the collapse of the dependent lung. The heart is displaced into the dependent hemithorax and is compressing the lung. The dependent aspect of the diaphragm is displaced cranially, also compressing the lung. There is reduced thoracic cage movement in the dependent hemithorax caused by compression. All of these factors contribute to the dependent atelectasis.

In addition to being misdiagnosed as lung disease, recumbent atelectasis is important because the increased radiographic opacity of the collapsed lung causes border effacement, or silhouetting, of any soft tissue lesion in the dependent lung, rendering it inconspicuous in the radiograph.¹ This phenomenon results commonly in lesions in the dependent lung not being visible radiographically in the lateral view (Fig. 25-2). The most significant aspect of this phenomenon is in the situation where the lesion is small and is not visible in the VD or DV view, such as commonly occurs with a lung nodule or small mass. If only one lateral view is made of such patients, and the lesion happens to be in the dependent lung and also in a region where it is not visible in the VD or DV radiograph, the lesion will be missed (Fig. 25-3). The collapse of the dependent lung is the main reason why both the left lateral and right lateral views should be included in every routine thoracic radiographic examination. If only one lateral view is obtained, a significant number of patients will be diagnosed incorrectly.²

Fig. 25-2 A, Ventrodorsal thoracic radiograph of a dog with pneumonia in the left cranial lobe. The increased opacity of this lobe is obvious. B, Right lateral radiograph. Multiple air bronchograms are visible because of the alveolar disease in the left cranial lobe *(white arrows).* The air bronchograms are highly conspicuous because the disease is in the nondependent lung and surrounded by aerated normal lung. C, Left lateral radiograph. The diseased left cranial lobe is dependent, and the entire left lung is atelectatic. The resultant increase in radiographic opacity of the left lung causes border effacement of the pneumonia, and it is not visible.

Given the amount of dependent atelectasis that occurs in lateral recumbency, it is not surprising that the overall radiographic opacity of the lung field will be increased in lateral versus VD or DV radiographs. Overinterpreting this increased opacity as lung disease is a common mistake (Fig. 25-4). If lung disease is suspected from one or both lateral views, it is critical that this be confirmed in the VD or DV view to rule out that the suspected lung disease is caused simply by atelectasis.

Fig. 25-4 Left lateral **(A)** and DV **(B)** thoracic radiographs of a dog, and close-up views of a region from the dorsocaudal lung field from the lateral radiograph **(C)** and the right-caudal lung field from the DV radiograph **(D).** The increased atelectasis that occurs in lateral views results in an overall increased opacity to the lungs (A and **C).** In the DV projection, where lungs are better aerated, they are more radiolucent (B and **D),** confirming that the opacity seen in the lateral view was an artifact of poor ventilation. The increased lung opacity commonly encountered in lateral views is often misinterpreted as lung disease. Any suspicious opacity seen in the lateral view must be confirmed as abnormal in either the VD or DV view.

The use of sedation or anesthesia for the radiographic examination will exacerbate the amount of recumbency-associated atelectasis that is present and will lead to an even greater increase in lung opacity that is misinterpreted easily as lung disease in lateral radiographs.

Other than conspicuity of lesions in the dependent versus the nondependent lung, there are some other important differences in the appearance of the lateral thoracic radiograph that can be attributed to whether the patient is in left or right lateral recumbency.

Cranial Lobe Vessels

Assessing the absolute and relative size of pulmonary vessels is an important process of thoracic radiographic interpretation.³ The pulmonary artery and vein in the right cranial lobe are used commonly as a basis for assessing the pulmonary circulation. It is important to be able to identify these two vessels specifically and to be able to compare the size of the right cranial lobe pulmonary artery to the size of the right cranial lobe pulmonary vein without having other vessels superimposed. The pulmonary vessels in the right cranial lung lobe are more conspicuous in a left lateral than in a right lateral radiograph (Fig. 25-5). This is caused mainly by the dorsal displacement of the right cranial lobe that occurs when the patient is in right lateral recumbency resulting in the right cranial lobar vessels being superimposed on the left cranial lobar vessels.