Dacryocystography can be helpful in evaluating patients with a history of unexplained epiphora or recurrent dacrocystitis. Following cannulation of the lacrimal puncta, contrast material is injected into the lacrimal duct to help delineate sites of obstruction or stenosis. Utilizing digital subtraction techniques, both static anatomy and the dynamics of lacrimal drainage can be assessed without obscuration by overlying osseous anatomy.5,6
Ultrasonography
Ultrasonography has long been used to evaluate tumors and other intraocular conditions. It has also been used in the evaluation and management of orbital disease. With the excellent resolution and easy accessibility of MRI and CT, at most centers orbital ultrasonography is typically used for specialized diagnostic and interventional purposes, rather than as a general modality for imaging the orbital contents. One reason for this is the lack of a skilled orbital ultrasonographer at many centers. Standard echography (contact B-scan and A-scan) is useful in the differentiation of various orbital vascular lesions.7 Similarly, color Doppler ultrasonography has been used as the sole imaging modality in a series of 20 patients with periocular and orbital infantile hemangioma.8
Several recent publications have demonstrated the value of orbital ultrasonography for specialized interventional applications.9–11 Orbital ultrasound guidance is utilized in the percutaneous drainage of macrocystic orbital lymphatic malformations, facilitating drainage of macrocysts followed by chemoablation (sclerotherapy).9 Doppler-guided core needle biopsy has been advocated, as has color Doppler-guided intralesional steroid injection into orbital infantile hemangiomas.10,11
Computed Tomography (CT Scan)
Since its invention in the 1970s, CT has revolutionized the practice of medicine, particularly in emergency care. Based on the principles of radiography, CT uses ionizing radiation to produce images in a noninvasive manner. CT images are generated when an x-ray tube and opposing detector rotate synchronously around a patient. Tissues absorb x-rays in differing amounts, depending on their molecular composition and density, and this difference in x-ray absorption is then used to generate cross-sectional images of the radiodensity of the internal structures. This radiodensity is measured in Hounsfield units. The Hounsfield unit scale ranges from −1000 to +1000, with water arbitrarily assigned the value of 0. Structures with higher attenuation relative to water have a positive value (soft tissues, bone, and calcification) and are displayed as bright pixels on CT images. Those structures that are relatively less radiodense than water have a negative value (air and fat) and are displayed as dark pixels on the resultant CT images.12
CT technology has evolved in three distinct aspects: image resolution, reduced radiation exposure, and decreasing scan acquisition time. Early generation axial scanners acquired images individually, one slice at a time, requiring scan times of several minutes. These examinations were susceptible to motion degradation and stairstep artifact if reconstructed outside the plane of acquisition.13 Newer-generation spiral/helical CT scanners acquire images continuously, resulting in reduced dose, decreased scanning time, and improved postprocessing capabilities.14 In addition, newer-generation scanners no longer use a single detector array, but rather many rows of multiple detectors. Detector arrays up to 320 rows are now in clinical use. Detectors of this size can image the head with one gantry rotation, nearly eliminating motion artifact. The isotropic images of modern scanners can be reformatted in any desired plane, which is particularly important in orbital imaging. Reconstructions of the orbit in the coronal, sagittal, and oblique sagittal planes supplements information acquired in the axial plane (Figs. 5.2 and 5.3). CT is the preferred imaging modality in several clinical situations including trauma (Figs. 5.4 and 5.5), suspected calcification (Fig. 5.6), and foreign body detection (Fig. 5.7).15 In the evaluation of an orbital mass, CT can offer additional information regarding osseous involvement that may be difficult to detect on MRI (Figs. 5.8 and 5.9). The detailed osseous anatomy available on CT is also helpful in assessing the thickness of cortical bone when planning orbital decompression. In cases of suspected orbital vascular malformations, a dedicated dynamic arterial phase CT angiogram and Valsalva-augmented venous phase CT protocol has been described. This protocol has the advantage of assessing arterial anatomy, lesion morphology, enhancement kinetics, and distensibility of orbital masses in one examination.16 The primary disadvantage of CT is radiation exposure, which is thought to increase the risk of cancer induction and cataract generation.
CT can be performed with or without the administration of intravenous contrast. The contrast used in CT imaging is iodine based and serves to identify blood vessels, structures that have a large blood volume, or areas with increased transit of contrast across the capillary endothelium or through the blood–brain barrier. These last situations are common in tumors and areas of inflammation, allowing contrast to highlight pathologic masses and to differentiate areas of inflammation or infection from surrounding normal tissues. The degree and pattern of contrast enhancement can often aid in identifying pathology and can enable the radiologist to generate a more tailored differential diagnosis. In orbital imaging, contrast is indicated in specific circumstances including assessment of infection or inflammation (Fig. 5.10), for characterizing an orbital mass (Fig. 5.11), or in the evaluation of a carotid cavernous fistula or other vascular pathology. Contrast is not typically required to evaluate trauma (Figs. 5.4 and 5.5), foreign bodies (Figs. 5.1, 5.7, and 5.12), or calcification (Fig. 5.6).17
CT Safety
There has been an increase in the awareness of the risks of radiation exposure, and an associated drive to limit the dose utilized to create CT images.18 In imaging the head, the most radiosensitive structure is the lens, and particular care is used to avoid unnecessary exposure of the lens.19 For orbital imaging, exposure of the lens is often unavoidable, and thus efforts are instead made to limit the overall dose of the examination. The relevant radiation dose depends on the imaging parameters utilized. In general, three factors influence CT dose: kilovoltage (kV), effective milliamperage (mA), and the physical length of the scan.20 The length of the scan must be appropriately prescribed to include only the relevant anatomy. Kilovoltage measures the energy of the photons generated by the x-ray tube. An increase in kilovoltage allows increased penetration of tissues, which can be diagnostically important in larger patients. However, dose increases exponentially with increasing kilovoltage, so kilovoltage should be set at the minimum level that allows an adequate number of x-rays to reach the detector. Effective milliamperage is a measure of the number of x-ray photons that pass through or are absorbed within a given volume of tissue. The change of dose with milliamperage is linear and typically automatically adjusted by the CT scanner to compensate for changes in tissue thickness as the patient proceeds through the CT scanner.
There is an important trade-off between section thickness and noise in CT imaging. If a slice is thinner, fewer x-ray photons have passed through the section of tissue being displayed, resulting in less sampling of the tissue and increased noise. For conventional filtered back projection reconstruction techniques, the noise increase is proportional to the square root of the change in image width.21 For example, if all other factors are held constant, a 2.5-mm thick image would have double the noise of a 10-mm thick image. These factors—noise, image width, and dose—must be appropriately balanced for each CT protocol.
National dose guidelines are published in several countries, typically as diagnostic reference levels set at the 75th percentile of reported doses for a particular type of examination.22,23 These normative data allow imaging centers to institute corrective action for protocols that fall outside the typical dose range. Continuous monitoring of protocols helps ensure that doses remain as low as reasonably achievable despite ongoing changes in CT techniques and equipment.
The radiation dose for a CT of the orbit is a fraction (currently approximately one-fifth) of the typical yearly background radiation dose.24 This CT dose is also less than the geographic variance in yearly background dose found within the United States.21 These low doses should not prevent obtaining clinically necessary information, particularly if there is no alternative modality able to provide these data.
Like any injected pharmaceutical, CT contrast has a small but important rate of adverse events. For modern low-osmolality contrast agents, the rate of true anaphylactic reaction is very small, estimated at 4 in 10,000 administrations (0.04%).25 A higher rate of atopic reactions and physiologic reactions (commonly vasovagal reactions) has been reported, with a frequency of 0.2% to 0.7%.25–27 For hives and other nonanaphylactic reactions, premedication before subsequent contrast-enhanced examinations with antihistamine agents and corticosteroid is a common practice.28 Of particular note, there is no evidence that patients with an allergy to seafood or shellfish have any greater risk of an allergic reaction to CT contrast compared with patients with other types of food allergies or asthma.29 Typically, food allergies and asthma are not considered a contraindication to CT contrast administration. It is also important to understand that CT and MRI contrast agents are chemically distinct; therefore, if an individual has an allergy to one type of contrast, those patients may be imaged with the alternative modality.
Contrast-induced nephropathy (CIN) has also been attributed to intravenous CT contrast administration, and screening of renal function is advocated for patients at risk for poor renal function (Table 5.1). This is of particular interest in patients with diabetes, as they have been reported to have higher rates of CIN following intravenous CT contrast administration. However, it should be noted that changes in renal function are common in patients following illness or hospital admission and is not always attributable to CIN. A recent review found that of eight studies of CIN that included a control group, only one found evidence for CIN.28 The evidence for setting an appropriate renal function cutoff level is also limited. Surveys of radiologists have shown that many groups choose a maximum serum creatinine cutoff between 1.5 mg/dL and 2.0 mg/dL, with lower values often used for patients with diabetes.30
Table 5.1
Patient Groups in Need of Creatinine Screening before Administration of Intravascular Iodinated Contrast
Age > 60 years
History of renal disease:
■ Dialysis
■ Kidney transplantation
■ Single kidney
■ Renal surgery
■ Renal cancer
History of hypertension requiring medical therapy
History of diabetes mellitus
Metformin or metformin-containing drugs*
