In imaging for pericardial effusion (see the images below), echocardiography and tomographic modalities (MRI, CT, EBT) are quite sensitive and can identify the presence of pericardial fluid even at the normal amount of 15-35 mL. Pericardial fluid is considered normal in the absence of pericardial disease if it appears as a homogeneous or echo-free space between visceral pericardium and parietal pericardium seen only during systole when the heart contracts inward, with less that 1 mm separation of the pericardial layers during diastole. Fat between the visceral and parietal layers may produce a false positive echo but is distinctive by MRI and CT. [1, 2, 3, 4, 5, 6, 7, 8]
Asymptomatic effusions are typically first detected by radiography performed for other reasons. A minimum of about 250 mL of fluid collection is required for detection through radiography that augments the cardiac silhouette. Increased pericardial fluid can be hydropericardium (transudate), true pericardial effusion (exudates), pyopericardium (if purulent), hemopericardium (in presence of blood), or mixtures of the above.
The normal pericardium is frequently identified on a lateral plain chest radiograph as a thin, linear opacity between the anterior subxiphoid mediastinal fat and subepicardial fat. In the posteroanterior (PA) view, the pericardium may be seen along the left heart border. [9, 10]
Pericardial effusion is characterized by accumulation of excess fluid in the pericardial space surrounding the heart. Most commonly, the fluid is exudative and results from pericardial injury or inflammation. Serosanguineous effusions are seen mainly in patients with tuberculous cancers, but they may also be encountered in uremic and viral disease or following mediastinal irradiation.
Hemopericardium is most commonly seen with trauma, myocardial rupture following myocardial infarction, myocardial or epicardial coronary artery rupture, catheter manipulation, aortic dissection with rupture into the pericardial space, or spontaneous hemorrhage in the presence of anticoagulant therapy. Chylopericardium is a rare condition that results from leakage or injury to the thoracic duct.
The presence of pericardial effusion generally indicates underlying pericardial disease; however, the clinical significance of pericardial effusion is mainly associated with its hemodynamic impact. The latter depends on the rate of fluid collection in the pericardial space, the rate of rise in the intrapericardial pressure, and resultant development of pericardial tamponade. A rapidly accumulating effusion, such as that associated with hemopericardium due to trauma, may result in tamponade with collection of as little as 100-200 mL of fluid, while a more gradual accumulation of fluid may allow for compensatory stretching of the pericardium and may not show tamponade, despite collection of fluid even in excess of 1500 mL.
Pericardiocentesis is needed in patients with hemodynamic compromise, tamponade, hemopericardium, or pyopericardium. However, in the absence of these factors, drainage is rarely indicated. Large effusions can sometimes be drained to relieve symptoms due to compression of surrounding lung and other structures. Pericardial drainage may occasionally be required to make a diagnosis based on examination of the pericardial fluid or pericardial biopsy samples.
Echocardiography is the most widely used imaging technique for the detection of pericardial effusion and/or thickening. A major advantage of echocardiography is its portability to the bedside to examine critically ill patients. The technique is noninvasive and is quite sensitive in imaging fluid-filled structures. When the pericardial fluid volume is small, it may appear as an anterior hypoechoic or echo-free space behind the left ventricle (LV), which could also represent a fat pad or a posterior or circumferential hypoechoic or echo-free space (the latter is most likely effusion). [11, 12]
When the pericardial effusion is large, the pericardial hypoechoic zone may expand to encircle the right ventricular (RV) apex. Rarely, echocardiography may be unable to identify pericardial fluid, especially in the presence of constriction, tumor, or hemorrhage. Echocardiography is also useful in assessing the hemodynamic impact of the effusion: right atrial inversion, right ventricle inversion, septal motion, and respiratory variation in Doppler transvalvular flow (>50% right, >25% left) indicate compromise.
If echocardiographic findings are inconclusive, CT or MRI can be helpful in detecting pericardial thickening, diffuse or loculated effusion, calcification, adjacent mediastinal and pulmonary disease, and neoplasm.
Electrocardiography (ECG) is of little or no diagnostic value. Sometimes, large pericardial effusions may exhibit electrical alternation of QRRS voltage or "pulsus alternans" reflecting free swing of the heart within the pericardial fluid with shifting electrical axis. Reduced ECG voltage is nonspecific. Nonetheless, massive effusions, such as those seen in severe myxedema, produce true "low" voltage that may parallel that seen with severe myocardial or hemodynamic abnormality. ECG may show diffuse PR depressions (often viewed as ST-T elevations), indicating pericardial or myocardial inflammation, diffuse T-wave inversions, and low voltage and suggesting effusion; the latter findings are not reliable. [13, 14]
MRI and CT
The extent of normal and abnormal pericardium is best appreciated with CT and MRI in most patients because of better resolution. With both CT and MRI, the anterior, lateral, and posterior portions of the pericardium are clearly separated from mediastinal fat. In addition, discontinuous areas of pericardial thickening and loculated effusions can also be identified. While the pericardial recesses are clearly defined by MRI and CT, they may on occasion mimic aortic dissection or mediastinal lymphadenopathy. [15, 16, 17]
MRI stripes can identify pericardial adhesion to underlying myocardium, and dynamic views of LV filling can identify constriction or restrictive disease. MRI strain maps can help confirm restrictive disease. MRI chemical sensitivity, by spatiospectral excitation and/or by inversion crossing, can distinguish fat from other materials in the pericardial space.
As pericardial fluid accumulates, the cardiac silhouette begins to enlarge and appears as flask-like, triangular, or globular in shape. The usual indentations and prominences normally seen along both the left and the right heart borders begin to diminish, such that the shape of the cardiac silhouette becomes globular and featureless.
Radiographic findings are only suggestive and often inconclusive except in patients with massive effusions. The degree of confidence is low, and false-positive and false-negative results are frequent. The shape of the cardiac silhouette on fluoroscopy or static images may raise the possibility of pericardial effusion but cannot decisively differentiate cardiomegaly from pericardial effusion.
Since the pericardium extends up to the pulmonary bifurcation, when a large pericardial effusion is present, the hilar structures are covered by the distended pericardial cavity. This appearance can help in differentiating a pericardial effusion from cardiomegaly, which does not hide the hilar vessels.
A "water-bottle" silhouette or unusually wide mediastinal shadow is suggestive of the presence of pericardial effusion, particularly when the lung fields are clear. Left pleural effusion is common with pericardial effusion, while bilateral pleural effusion is seen more often with congestive heart failure. On a well-penetrated lateral chest radiograph, the presence of pericardial fluid is suggested by radiolucent fat lines within the cardiopericardial silhouette. The retrosternal space is narrowed by the enlarging cardiac shadow.
When pericardial effusion is present, the subepicardial fat is displaced posteriorly by the higher-density fluid. In this situation, a wide opaque vertical band may be seen between the anterior border of the heart and the mediastinum, called the "epicardial fat pad sign".
A CT scan provides a distinct image of the pericardium in most patients. Differentiation of the pericardial line from the myocardium is enabled by the presence of a small amount of epicardial and pericardial fat. This fat may be visible on CT scans in over 95% of instances. Although CT is extremely sensitive for the detection of pericardial effusion, it has some limitations in defining loculated effusions, hemorrhagic effusions, and especially pericardial thickening. Thus, it is complementary to echocardiography in the diagnosis and assessment of pericardial disease. Also, unless gated imaging is performed, CT does not assess hemodynamic compromise. CT is not as good as MRI in evaluating possible restrictive disease.
CT also frequently demonstrates the superior recesses of the pericardium extending over the ascending aorta and lateral to the main pulmonary artery. These recesses may be distended in the presence of a pericardial effusion. The attenuation of the effusion on CT may be helpful in suggesting the etiology. Attenuation greater than water suggests malignancy, blood, pus, or effusion secondary to hypothyroidism. [15, 16]
Magnetic Resonance Imaging
MRI can detect pericardial effusion with clarity. The degree of confidence is high, and the rate of false-positive and false-negative results is low. In spin-echo images, the pericardial effusion or portions of it may appear as a signal void (dark) sac because of moving fluid in the pericardial cavity. However, the pericardial effusion is bright on MRIs obtained with a gradient-echo sequence. Hemorrhagic effusions have the opposite appearance; they have high signal intensity on T1-weighted spin echo images and low intensity on gradient echo images. Intrapericardial masses, cysts, and diffuse thickening are generally well demonstrated with MRI imaging.
MRI can clearly depict pericardial recesses, mediastinal fat, and other similar anatomic structures within the pericardial sac. Dark-blood (double inversion recovery) imaging can measure pericardial thickness accurately. Fat-sensitive imaging (based on chemical shift or on T1 values) can discern pericardial fat from other materials. Dynamic MRI can identify constrictive disease (failure to expand for the last 6/20 frames of diastole). MRI can also assess for adhesions by placing demagnetization stripes to observe for translational motion at the pericardium, and it can assess for restrictive disease by strain mapping. 
In the absence of any significant pericardial effusion, echocardiography in normal persons shows small amounts of fluid mostly occupying posterior space between the LV wall and the parietal pericardium. This fluid is usually visible only during systole when the LV wall moves inward away from the pericardium. As the amount of fluid increases, the posterior space begins to show it in both systole and diastole. As effusions become even larger, fluid appears anterior to the RV, as well. Finally, with the development of tamponade, the fluid can be found posterior to the left atrium (in the pericardial oblique sinus).
Echocardiography can provide an estimate of the size of effusions. Generally, small effusions cause an echo-free space in systole and diastole of less than 10 mm; moderate effusions, 10-20 mm; and large effusions, greater than 20 mm. The size of pericardial effusion is a powerful predictor of overall prognosis. For example, large effusions generally indicate more serious disease.
If only an anterior echo-free space is seen, posterior adhesions might be present, although epicardial fat is usually more likely. Rarely, an infiltrative lesion, often malignant, can cause anterior echo-lucent space. Epicardial fat, when present posteriorly or circumferentially, may have to be differentiated by CT or MRI scan.
The degree of confidence is high, and false-positive and false-negative findings are infrequent. Pericardial fat may produce a false-positive echo appearance of effusion. Pleural effusions may sometimes be mistaken for pericardial when views are poor or limited.
Transesophageal echocardiography (TEE) is superior to transthoracic echocardiography (TTE), especially for identifying metastases, pericardial thickening, and clots, but it tends to cause underestimation of the volume and distribution of effusions. Three-dimensional echocardiography is a promising new technology. Although echocardiography is sufficient in most cases, CT and MRI may be needed if echocardiographic results are equivocal. 
Echocardiographic findings in pericardial effusion may include the following:
Echo-free space: (1) posterior to LV (small-to-moderate effusion); (2) posterior and anterior (moderate-to-large effusion); (3) behind left atrium (large-to-very large effusion and/or anterior adhesion
Diminished mobility of posterior pericardium-to-lung interface
Enhanced RV wall mobility unmasked in presence of anterior fluid
"Swinging" heart (large effusions, usually tamponade): (1) RV and LV walls move in synchrony; (2) periodicity 1:1 or 2:1 (one or two swings per cardiac cycle); (3)2:1 swing is characteristic of cardiac tamponade; (4) pseudoparadoxic motion of LV posterior wall; (5) mitral and/or tricuspid pseudoprolapse; (6) mitral systolic anterior motion; (7) alternating mitral E-F slope and aortic opening excursion; (8) aortic valve exhibiting midsystolic closure; (9) pulmonic valve with midsystolic notch
Hemopericardium with blood clots identifiable by echocardiography
Inspiratory decrease in LV ejection time (with effusion)
Echocardiographic findings in cardiac tamponade may include the following:
RV compression: RV diameters are reduced, particularly RV outflow tract (to 7 mm or less); early diastolic collapse of RV may be seen
RA free-wall collapse during late diastole
RA isovolumic contraction is prolonged to occupy one third of the cardiac cycle
LA free-wall compression in patients with fluid posterior to the left atrium
LV free-wall may exhibit paradoxical movement
Superior vena cava (SVC) and inferior vena cava (IVC) may show congestion (unless patient is relatively volume depleted)
IVC is usually greater than 2.2 cm in diameter with less than 50% inspiratory compression
Exaggerated inspiratory effects, especially with pulsus paradoxus, may include RV expansion, interventricular septum shift to the left, and LV compression
Mitral changes, with reduced D-E amplitude or E-F slope and delayed mitral opening time
Aortic valve with premature closure
Echocardiographic stroke volume diminished
RV epicardial notching during isovolumic contraction
Course vibrations of LV posterior wall
Pseudohypertrophy or apparent wall thickening due to compression
Doppler studies may include the following:
Generally reduced flows and stroke volumes
Exaggerated inspiratory augmentation of right-sided flows and reduction in left-sided flows (eg, >50% inspiratory change in peak transvalvular Doppler of right-sided valves, >25% at left-sided valves)
Respiratory variation in superior and inferior vena caval flow velocities, particularly marked in tamponade
Reduced expiratory diastolic SVC flow
Hepatic vein expiratory effects: marked atrial flow reversal (AR wave), marked decrease or reversal of diastolic forward flow, occasional systolic flow reversal
On transesophageal echocardiograms (TEE), expiratory increase in pulmonary vein diastolic forward flow can be appreciated
Marked inspiratory decrease in LV ejection time occurs, along with increased RV ejection time
Marked inspiratory increase in LV isovolumic relaxation time is encountered, in association with decreased RV isovolumic relaxation time
The degree of confidence is low.
It may be necessary to confirm the diagnosis of tamponade by right-heart catheterization or full cardiac catheterization and angiography, particularly when study of the heart and coronary arteries becomes necessary. Generally, hemodynamic measurements during cardiac catheterization show high pressures throughout ventricular diastole and near-equilibration (>5 mm Hg) of atrial and ventricular diastolic pressures. The degree of confidence is good. False-positive and false-negative findings are rare.
Most tamponades equilibrate at 16-20 mm Hg, while some patients may equilibriate at 6-12 mm Hg with low-pressure tamponade. Atrial hemodynamic tracings show absent or reduced 'y' descent. Arterial and pericardial catheters disclose exaggerated respiratory pressure fluctuations. Notably, despite high central filling pressures, cardiac tamponade does not cause alveolar pulmonary edema.