Canine Thoracic Radiographs: Projections and Key Findings

Canine thoracic radiographs remain one of the most commonly performed diagnostic imaging studies in veterinary medicine, offering a fast, non-invasive look at the heart, lungs, airways, mediastinum, and surrounding structures. Getting the most out of these images depends on choosing the right projections for the clinical question and knowing which anatomical landmarks and patterns to evaluate on each view. The subtleties matter more than many clinicians realize, and the difference between a two-view and three-view study, or between a dorsoventral and ventrodorsal projection, can change a diagnosis in a meaningful percentage of cases.

Standard Projections and When Each One Matters

A thoracic radiographic study in dogs typically includes at least two views: a lateral and an orthogonal view (either ventrodorsal or dorsoventral). The lateral projection is the workhorse. Most dogs are positioned in right lateral recumbency for the standard lateral view, but there are real diagnostic consequences to which side is down. Research has shown that the right cranial lung lobe is better aerated when a dog lies in left lateral recumbency, and lesions in certain lobes can be masked when the affected lobe is the dependent (down) side or enhanced when it is the non-dependent (up) side. This happens because the dependent lung gets compressed under the weight of the mediastinum and the opposite lung, while the uppermost lung inflates more fully.¹ In practice, this means that if you suspect disease in a specific lung lobe, placing that lobe non-dependent (up) can make the lesion easier to see.

The choice between a dorsoventral (DV) and ventrodorsal (VV) projection is not just a matter of patient comfort. The DV view, where the dog lies on its sternum, is favored for evaluating the cardiac silhouette in a more consistent position, for assessing pulmonary lobar vessels, and for examining structures in the dorsal thorax such as the hilar lymph nodes, caudal dorsal lung fields, trachea, mainstem bronchi, and the left atrium.² The VV view, taken with the dog on its back, spreads the lung fields more symmetrically and is often preferred for evaluating lung pathology across both sides. Critically ill or dyspneic patients are usually positioned DV because lying on their backs can compromise breathing.

Why Three Views Beat Two for Detecting Lung Nodules

When the clinical question involves screening for pulmonary metastatic disease, the number of radiographic views matters substantially. A standard two-view study (one lateral plus one VD or DV) can miss nodules that a three-view study (right lateral, left lateral, and VD or DV) would catch. One study comparing two-view and three-view protocols found that dropping any single view from a three-view study would change the diagnosis in roughly 12 to 15 percent of patients with structured interstitial pulmonary disease, including metastatic lesions.³ That is a large enough proportion to justify the extra view whenever metastasis is a concern.

Separate work on radiographic sensitivity for lung metastases has reinforced this, recommending that the right lateral recumbent view be included in any diagnostic protocol for detecting pulmonary metastases. The same study also recommended using a minimum of two independent readers, or defaulting to a three-view protocol when only one reader is available, to compensate for the inherent limits of human pattern recognition on chest films.⁴ The practical takeaway is straightforward: if you are staging a cancer patient, take all three views.

Measuring Heart Size with the Vertebral Heart Score

Subjective assessment of the cardiac silhouette on a lateral radiograph is unreliable across different observers. The vertebral heart score (VHS), described by Buchanan and Bücheler, addressed this by providing a quantitative, reproducible method. The technique involves measuring the long and short axes of the heart on a lateral radiograph, then comparing those combined lengths against the vertebral bodies starting at the cranial edge of T4.⁵ A normal VHS in most breeds falls around 9.7 vertebral units, though breed-specific reference ranges exist because deep-chested and barrel-chested breeds have different normal silhouettes.

The method has proven remarkably consistent across experienced interpreters. When nine veterinarians measured VHS on radiographs from one normal dog and one dog with severe cardiomegaly, the coefficients of variation were only about 2.7 and 2.8 percent, respectively.⁶ That level of agreement means the VHS is useful not just as a one-time measurement but also for tracking changes over serial radiographs in patients with progressive heart disease.

Vertebral Left Atrial Size

While VHS captures overall cardiac enlargement, left atrial size is often the specific chamber of interest in dogs with mitral valve disease. The vertebral left atrial size (VLAS) method was developed to quantify left atrial dimensions on the same right lateral radiograph used for VHS. Validation studies in healthy dogs found excellent intra-observer and inter-observer agreement for VLAS, with intraclass correlation coefficients at or above 0.91.⁷ VLAS has become increasingly relevant in clinical cardiology because the staging of myxomatous mitral valve disease (the most common heart disease in dogs) hinges partly on whether left atrial enlargement has developed. Radiographs are more accessible than echocardiography in many general practices, so a reliable radiographic metric for left atrial size fills a genuine gap.

Reading the Pulmonary Vessels

The pulmonary arteries and veins visible on a thoracic radiograph carry diagnostic information about both cardiac and pulmonary conditions. In dogs with mitral regurgitation, the pulmonary veins become engorged because blood backs up from the left atrium into the pulmonary circulation. One approach compares the diameter of the right caudal pulmonary vein to its accompanying pulmonary artery. Research in dogs with mitral regurgitation found that this comparative method had about 80 percent sensitivity and roughly 83 percent specificity for detecting the disease. The study recommended using the accompanying pulmonary artery diameter and a threshold of 1.22 times the width of the ninth rib as radiographic criteria for distinguishing healthy dogs from those with mitral regurgitation.⁸

Pulmonary artery enlargement, on the other hand, points toward conditions that increase right-sided pressures, such as pulmonary hypertension or heartworm disease. Veins that are larger than their paired arteries suggest left-sided congestion; arteries bigger than their paired veins suggest right-sided or pulmonary vascular disease. This artery-versus-vein comparison is one of the most clinically useful assessments on a thoracic radiograph, and it can be performed quickly once you know the landmarks.

Lung Patterns and Their Diagnostic Meaning

Radiographic lung patterns are broadly categorized as alveolar, interstitial (structured or unstructured), bronchial, or vascular. In practice, patterns frequently overlap, especially in sick patients. Noncardiogenic pulmonary edema, for instance, tends to produce a mixed interstitial-to-alveolar pattern. One retrospective study of 60 animals with noncardiogenic pulmonary edema found that about three-quarters showed a mixed pattern, and the distribution was bilateral in a similar proportion. The infiltrates were peripheral in over 90 percent of cases and dorsal in more than half.⁹ When edema was unilateral, the right lung lobes were affected in the large majority of cases.

A separate, smaller retrospective study confirmed that the most common radiographic presentation of noncardiogenic pulmonary edema was a diffuse mixed interstitial-to-alveolar pattern, seen in just over half of cases.¹⁰ The dorsal and peripheral distribution of noncardiogenic edema contrasts with the perihilar and ventral distribution more typical of cardiogenic pulmonary edema, and that distinction is a key differentiator on radiographs. If a dog presents with respiratory distress and the edema pattern is dorsally distributed and peripheral, the clinician should think about causes other than heart failure, such as upper airway obstruction, seizures, electrocution, or neurologic events.

Postobstructive pulmonary edema, which develops after relief of an upper airway obstruction, tends to look different even within the noncardiogenic category. It is more often asymmetric, unilateral, and dorsal compared to other forms of noncardiogenic edema.⁹ Recognizing this subtype matters because it usually resolves with supportive care and does not require the same treatment as heart failure-related edema.

Evaluating the Trachea for Collapse

Tracheal collapse is one of the most common airway disorders in small-breed dogs, and thoracic radiographs are usually the first imaging tool used to assess it. Standard radiographs capture a single moment in the respiratory cycle, so the degree of collapse seen depends on timing. In healthy small-breed dogs, the average variation in tracheal height between maximum inspiration and maximum expiration is small, under about 6 percent in most individuals, though it can range as high as 21 percent in some healthy dogs.¹¹ Inspiratory and expiratory radiographs taken in right lateral recumbency provide an accurate measure of tracheal height and can serve as a practical alternative to fluoroscopy in many cases.

That said, fluoroscopy does outperform static radiographs in some regions of the airway. A direct comparison found no significant difference between radiography and fluoroscopy for measuring tracheal diameter change in most segments, but at the carina, fluoroscopy detected a significantly greater degree of collapse than radiographs did.¹² The carina is a transition zone where the trachea meets the mainstem bronchi, and dynamic imaging is particularly valuable there because the collapse pattern can be different during different phases of breathing. If radiographs are inconclusive in a dog with a classic honking cough, fluoroscopy is the logical next step.

Mediastinal Findings

The mediastinum, the central compartment of the thorax between the two lung fields, is often overlooked on canine thoracic radiographs because it does not stand out unless something is wrong. One of the more dramatic mediastinal findings is pneumomediastinum, or free air in the mediastinal space. In a case series of seventeen dogs with pneumomediastinum, the most frequent and reliable radiographic signs were visualization of both the inner and outer walls of the trachea (the tracheal wall sign), identification of the azygos vein outlined by air, and the presence of discrete mediastinal air pockets.¹³ Air in the mediastinum can track in from a ruptured trachea, torn esophagus, or extension from subcutaneous or retroperitoneal emphysema, so recognizing these signs often prompts a search for the underlying cause.

Mediastinal masses, including lymphoma and thymoma, typically present as a widened mediastinum on the VD or DV view and a soft tissue opacity in the cranial thorax on the lateral. The cranial mediastinum in young dogs can normally appear wide due to the thymus, so age and clinical context matter when interpreting these findings.

The Esophagus and Diaphragm

The esophagus is not routinely visible on thoracic radiographs unless it is dilated with air, fluid, or food. When it is visibly distended, the condition is called megaesophagus, and the radiographic finding is a tubular soft-tissue or gas opacity running dorsal to the trachea and heart. Numerous underlying causes exist, including myasthenia gravis, which is one of the more treatable ones. Researchers attempted to use the relative diameter of the esophagus on radiographs to distinguish myasthenia gravis from other causes of megaesophagus but found that although there was a small statistically significant difference in median relative esophageal diameter between the two groups, the ranges overlapped completely. The measurement was deemed to have limited diagnostic utility for that purpose.¹⁴ In practice, an acetylcholine receptor antibody titer remains the definitive test for myasthenia gravis; the radiograph’s role is to confirm the megaesophagus itself and to check for aspiration pneumonia, which commonly complicates it.

The diaphragm is the caudal boundary of the thorax and a structure that deserves careful scrutiny. Diaphragmatic hernia, whether traumatic or congenital, allows abdominal organs to shift into the thorax. A review of 60 cases in dogs and cats found that at least two radiographic views at different angles were necessary for a valid diagnosis, because some signs were visible only on certain projections. Beyond the classic findings of loss of the diaphragmatic silhouette and soft-tissue opacity in the caudal thorax, useful radiographic signs included the position of the stomach axis and displacement of tracheal and bronchial segments.¹⁵ A single-view study that looks normal does not rule out a diaphragmatic hernia, particularly in a trauma patient.

Sternal and Other Extrathoracic Structures

The sternum is not the reason most thoracic radiographs are taken, but it is visible on every lateral view and abnormalities are fairly common. Sternal assessment on thoracic radiographs involves evaluating the alignment, number, shape, and opacity of the sternebrae along with the width and opacity of the intersternebral cartilages.¹⁶ Many sternal variations are incidental and clinically insignificant, such as congenitally fused or malformed sternebrae. But pathologic changes, including lytic or proliferative lesions, can indicate neoplasia, infection, or metabolic disease and should not be dismissed as normal variants without consideration of the clinical picture.

Other extrathoracic structures visible on thoracic radiographs include the thoracic spine (where disk disease, vertebral tumors, or discospondylitis might be spotted incidentally), the cranial abdomen (a gas-filled stomach or hepatomegaly can be suggestive), and the soft tissues of the thoracic wall (subcutaneous masses or emphysema). A systematic approach to reading thoracic radiographs includes checking all of these peripheral structures so that clinically relevant findings outside the lungs and heart are not missed.

When Radiographs Are Not Enough

Thoracic radiographs are highly accessible, relatively inexpensive, and remarkably informative for a wide range of conditions. But they have real limits, and computed tomography (CT) clearly surpasses them for certain questions. The starkest contrast appears in the detection of pulmonary nodules in cancer patients. One study found that only about 9 percent of nodules detected by CT were identified on thoracic radiographs. The lower size threshold for detecting a nodule was approximately 1 mm on CT versus 7 to 9 mm on radiographs.¹⁷ Nodules were also detected in significantly more lung lobes on CT than on radiographs, meaning that radiographs not only miss small nodules but also underestimate the distribution of disease.

A separate comparison of three-view radiography versus CT found that radiographic sensitivity for pulmonary nodules ranged from about 71 to 95 percent depending on the reader, with specificity between 67 and 92 percent. CT was consistently more sensitive.¹⁸ The clinical implication is clear: when accurate staging of metastatic disease would change the treatment plan, CT is the superior modality. Radiographs are still a reasonable first-line screen, but a clean set of chest films in a cancer patient does not rule out pulmonary metastasis, especially for nodules under a centimeter. This is particularly relevant for tumors with high metastatic potential, such as osteosarcoma or hemangiosarcoma, where the difference between detecting two nodules versus twenty changes whether surgery or chemotherapy is pursued.

Image Quality and Common Technical Pitfalls

Even the most experienced radiologist cannot interpret a poor-quality image, and thoracic radiographs are especially unforgiving of technical errors because the lungs, heart, and soft tissues span a huge range of tissue density on a single image. Motion artifacts are among the most common problems, causing evident distortion of anatomical structures that can obscure subtle lesions or mimic pathology that is not actually present.¹⁹ Exposure timing during respiration matters, too: radiographs taken at peak inspiration show the lungs at their most inflated and provide the best contrast between air-filled lung and soft-tissue or fluid opacities. An expiratory film can make normal lungs look hazy and suggest disease where none exists.

Rotation of the patient, incomplete inclusion of the thorax on the image, and improper exposure settings all degrade diagnostic quality. On a well-positioned lateral view, the rib heads should overlap and the diaphragmatic crura should be clearly separated. On a VD or DV view, the spinous processes should bisect the sternum symmetrically. When these landmarks are off, measurements like VHS become unreliable and apparent cardiac enlargement could simply be an artifact of positioning. For any thoracic radiograph that looks abnormal but does not fit the clinical picture, re-taking the image with better technique is often more productive than chasing a diagnosis that might not exist.