Chapter 84 – Orbital Imaging Techniques
JONATHAN J. DUTTON
• Computed tomography: An imaging technique where contrast differences are based on tissue density based on the passage of X-rays through tissues.
• Magnetic resonance imaging: An imaging technique where density differences are based on tissue proton density and their resonance characteristics based on biochemical relationships within the atomic lattice.
• Orbital echography: An imaging system where density differences are based on tissue characteristics reflecting the passage, reflection, and refraction of sound waves through the tissue.
Radiographic examination is an essential step in the evaluation of all patients who have suspected orbital disease. This routine frequently contributes to a specific diagnosis and also may help the physician plan the most appropriate medical therapy or surgical approach. Computed tomography and magnetic resonance imaging have largely replaced older techniques. Orbital echography is another valuable tool that can provide information not obtained easily with other techniques.
In computed tomography (CT), thin, collimated X-ray beams pass through tissue along the rows and columns of an intersecting matrix. The area defined by any two intersecting beams is referred to as a pixel and is analogous to a single dot in a newspaper photograph. Since the X-ray beam has a certain thickness, the area of the beam intersection actually defines a volumetric space, referred to as the voxel. As X-ray beams traverse the body, they are weakened or attenuated according to the density of the tissues through which they pass ( Fig. 84-1 ). The degree of attenuation of any two intersecting beams that emerge from a volume of tissue allows the calculation of the mean attenuation value for all the tissues included within the area of intersection of the beams, or voxel. The smaller the pixel and the thinner the tissue slice, the smaller the volume of the voxel and the higher the resolution of the final image.
Each voxel is assigned an average attenuation value by the computer based on the mean attenuation of the X-rays that pass through the voxel. These values are designated in Hounsfield units, a 2000-unit scale from -1000 to +1000. By arbitrary convention, the density of air is assigned a value of -1000, the density of water is 0, and the density of bone is +1000. For visualization by the human eye, this scale is reduced to 32–64 gray levels between black and white on the radiographic film. Thus, air appears black on the film, and bone appears white. A density greater than that of bone, such as that of a metallic foreign body, also appears white.
For specific anatomical detail, the image is manipulated by setting “windows.” The window level refers to the Hounsfield unit on which a narrow range of units is to be centered. The window range is the inclusive number of Hounsfield
Figure 84-1 Computed tomography. Intersecting matrix of collimated X-ray beams passes through tissue and is attenuated according to differences in tissue density.
units above and below this level that are to be expanded into the black-to-white scale for imaging. In the examination of a soft tissue lesion such as a hemangioma, for example, the window level may be set to +50, the density of muscle, and the window range to plus and minus 200 units. With these window settings, -150 on the scale appears black, and +250 appears white. All attenuation values below -150 also appear black, so there is no detail visible in the orbital fat. Also, all those values above +250 appear white, so there is no detail seen in bone. Similarly, bone windows are centered on about +800 for visualization of bone details ( Fig. 84-2 ).
Orbital CT routinely should include scans in both the axial and the coronal planes. Contrast enhancement is generally less useful than for brain studies because of the lack of a blood-orbit barrier, but it may provide valuable information about the nature of particular types of lesions. Unless contraindicated, contrast studies should be included in all orbital scans. Surface coil technology and fat-suppression algorithms may significantly increase resolution and contrast.    
MAGNETIC RESONANCE IMAGING
The technique of magnetic resonance imaging (MRI) offers several advantages over CT.    Because low-resonance signal is generated by bone, soft tissue visualization in the region of the orbital apex, optic canal, and cavernous sinus is not as degraded as in CT scans.    Manipulation of resonance signals provides contrast variability and tissue differentiation unobtainable with any X-ray technique. Surface coil technology, improvements in signal-to-noise ratios, and techniques to suppress the high fat signal on T1-weighted images have improved visualization of orbital lesions.     For intracranial lesions in the vicinity of the optic chiasm and suprasellar cistern, MRI has proved especially valuable and has all but replaced older techniques.
The generation of a magnetic resonance signal depends on the presence of magnetic isotopes of common elements in biological
Figure 84-2 Axial computed tomography of the orbits. A, Tissue density window. B, Bone density window.
Figure 84-3 Spinning proton nuclei. The left diagram depicts random orientation found in normal tissues with no set magnetic field. The right diagram depicts orientation with a net magnetic moment aligned with an external magnetic field.
tissues. The isotope most frequently imaged is the ubiquitous hydrogen nucleus or proton.  This proton is normally in a state of axial spin. This spinning charged particle generates a magnetic field, with north and south poles, analogous to a bar magnet. Under normal conditions, all the nuclei in a given volume of tissue are oriented randomly, with no net magnetic vector. When placed within a strong external magnetic field, the individual
Figure 84-4 Orientation of spinning nuclei. In the top diagram, all spinning nuclei are distributed evenly around the magnetic moment. In the bottom diagram, each spinning nucleus exhibits a wobble of its axis with one end (*) precessing around the direction of the mean magnetic moment.
Figure 84-5 Radiofrequency signal. With exposure to an external radiofrequency (RF) signal at the Larmor frequency, the spinning nuclear axes tip away from the direction of the magnetic field and group to one side.
protons align with the direction of the external magnetic field, either parallel or antiparallel ( Fig. 84-3 ). Because a slight preponderance of alignments is parallel to the direction of the magnetic field, the tissue assumes a mean magnetic moment in the same orientation.  Most of the axes of individual protons are not aligned perfectly with the direction of the magnetic field, but lie at various small angles to the mean magnetic moment. They are distributed equally through the 360° around the magnetic moment like spinning tops, and these inclined axes wobble, with one pole remaining stationary and the other revolving, or “precessing,” around the direction of the mean magnetic moment ( Fig. 84-4 ). The rotating axes, therefore, describe a conical surface. The angular velocity of precession is determined by the strength of the external magnetic field and by an intrinsic property of the particular atomic nucleus. This is called the gyromagnetic ratio, which is proportional to the magnetic moment. The relationship between these parameters is defined by the Larmor equation, and the resultant angular velocity is the resonant or Larmor frequency.
When this system is exposed to an external radiofrequency (RF) pulse at the Larmor frequency, energy is absorbed by the atomic nuclei. As the spinning nuclei move to higher energy levels, the angular orientation of their axes to the direction of the external magnetic field increases. Also, an induced magnetic field perpendicular to the RF pulse direction realigns the individual atomic axes to one side of the direction of the external magnetic field ( Fig. 84-5 ). When the RF signal is turned off, the spinning nuclei return to equilibrium by giving up energy to the environment, again at the Larmor frequency. Return to equilibrium occurs by two simultaneous decay, or relaxation, processes—T1 and T2.
During T1 relaxation, the individual nuclear axes realign parallel to the direction of the external magnetic field. In the process, they give up their absorbed energy. The time required for this process is the T1, or spin-lattice, relaxation time. It is influenced by the interaction of the proton with other atoms within the molecular lattice, by temperature, and by viscosity of the tissue. A high T1 relaxation time yields maximal energy release per unit time and, therefore, a higher resonance signal and brighter image.
During the RF pulse signal, while the atomic nuclei are still grouped on one side of the mean magnetic axis, they generate a resonant signal. This results from the tipped net magnetic vector of the spinning protons constantly cutting across the lines of force of the external magnetic field, thus generating a small alternating-current voltage. After the RF pulse is stopped, the atomic nuclei redistribute themselves evenly 360° around the direction of the external magnetic field; as they do so, the strength of this signal decreases because of canceling vectors. The time for complete decay of this resonant signal is the T2, or spin–spin relaxation time. It is influenced by the induced magnetic fields generated around adjacent spinning nuclei. As with the T1 times, biochemical differences between tissues confer slightly different T2 relaxation times to the protons in the tissue. Since the T1 and T2 signal strengths determine the contrast intensity, these biochemical differences result in contrast differentiation on the final image. Since small differences in T1 and T2 relaxation can be detected easily, contrast differentiation between adjacent tissues on MRI is considerably better than that with CT.
The T1 and T2 signals are measured by RF detectors. They detect in mass fashion all similar signals at the Larmor frequency, regardless of their specific location within the tissue. Spatial encoding of resonant signals from particular small blocks of tissue is necessary to create a visually meaningful two-dimensional image. This is achieved by deformation of the external magnetic field using gradient coils, such that the protons in every small volume of examined tissue (voxel) have a unique magnetic field strength and, therefore, a unique Larmor frequency. The detected Larmor frequency identifies the precise location of the signal, and thereby a topographical image can be created. The introduction of surface receiver coils placed close to the region of study has improved detection efficiency and the signal-to-noise ratio, thus decreasing the influence of surrounding magnetic aberrations.
Normal Orbital Anatomy in the Axial Plane
AXIAL SECTION THROUGH THE LOWERMOST ORBIT.
The orbital floor appears as a thin, oblique density that runs from anteromedial to posterolateral, separating the orbit from the maxillary sinus.   Since the floor gradually slopes backward and upward, successively higher cross sections are cut in axial scan sequences. The orbital cavity is bounded medially by the anterior lacrimal crest and laterally by the lateral rim of the zygomatic bone. Posterior to the orbit is the cranial base. A thin line arches across the orbital opening from the medial to the lateral bony rims; this represents the lower eyelid and orbital septum.
Depending on the level of the cut, the orbital cavity may appear empty (because it contains only orbital fat), or it may contain a rounded density that represents the sclera cut tangentially. Occasionally, the inferior oblique muscle is seen as an oblique band of medium-density tissue, and in slightly higher sections, the inferior rectus muscle may appear as a density adjacent to the globe posteriorly.
AXIAL SECTION THROUGH THE INFERIOR ORBIT.
In low axial sections through the inferior orbit, the floor again appears as a thin density that separates the orbital cavity from the maxillary sinus. In the posterolateral corner of the orbit, where the floor approaches the lateral wall, a channel separates the body of the sphenoid bone from the greater wing. This is the inferior orbital fissure. Posteriorly, behind the inferior rectus muscle, this fissure
Figure 84-6 Axial CT scan through the midorbit. The globe and optic nerve are seen in the axial plane, along with the medial and lateral rectus muscles. E, Ethmoid sinus; LR, lateral rectus; MR, medial rectus; ON, optic nerve; S, sphenoid sinus.
Figure 84-7 Axial MRI scan through the midorbit. The section is slightly higher than in Figure 84-6 . The optic nerves are seen passing back to the optic chiasm. E, Ethmoid sinus; LR, lateral rectus; MR, medial rectus; ON, optic nerve; S, sphenoid sinus.
communicates between the orbital space and the pterygopalatine fossa. In the anteromedial corner of the orbit, the lacrimal sac fossa is seen as a depression in the orbital process of the maxillary bone.
Within the orbital space, a central rounded density represents the globe. Since the vitreous is primarily aqueous, it appears empty (black) on CT. However, on MRI scans, the vitreous appears dark on T1-weighted sequences and bright on T2-weighted sequences. Just posterior to the globe is a rounded density that lies on the midportion of the orbital floor and is discontinuous with the globe. This is the inferior rectus muscle cut in cross section.
AXIAL SECTION THROUGH THE MIDORBIT.
On axial scans through the midorbit, the globe is seen in equatorial section ( Figs. 84-6 and 84-7 ). Anteriorly, the lens is seen as an oval density. On MRI sections, the ciliary body can be distinguished on either side of the lens. Behind the globe, the optic nerve is seen to emerge from the posterior sclera and run toward the orbital apex.
In the midorbit, a gently curved enhancing line crosses the orbit from lateral to medial. This is the superior ophthalmic vein. Near the orbital apex, a small enhancing vessel is seen
to cross over the optic nerve from lateral to medial. This is the second portion of the ophthalmic artery. Along the orbital walls are the lateral and medial rectus muscles. At slightly higher levels, both the medial rectus and superior oblique muscles are often seen together. On either side of the midline are the ethmoid sinuses, with the thin lamina papyracea that forms the medial orbital wall. Just medial to the laminae are the ethmoid air cells.
AXIAL SECTION THROUGH THE SUPERIOR ORBIT.
At this level, the orbital contour is narrower and terminates posteriorly in a rounded angle above the level of the optic canals. Within the orbital outline, the globe is represented in cross section above the level of the lens. Along the medial wall is the superior oblique muscle that passes through the trochlea anteromedially. Near the orbital apex, the superior rectus muscle appears as a broad band of tissue directed toward the globe. The superior ophthalmic vein is seen as a curvilinear enhancing structure that crosses from anteromedial to posterolateral just below the muscle. Anterolaterally, near the lateral orbital rim, the lacrimal gland appears as an oval density seen between the zygomatic bone and the globe.
AXIAL SECTION THROUGH THE ORBITAL ROOF.
In axial sections above the level of the globe, the orbit appears as a rounded contour posteriorly. Since the roof lies in a plane oblique to the tissue slice, in each higher section the roof lies progressively more anteriorly as it approaches the orbital rim. The levator muscle is seen as a broad band that extends from the superior orbital rim backward along the roof. Anteromedially, the trochlea is seen clearly, and at appropriate levels the superior oblique tendon can be visualized as it fans out over the superior globe below the superior rectus muscle insertion.
Normal Orbital Anatomy in the Coronal Plane
CORONAL SECTION THROUGH THE ANTERIORMOST ORBIT.
In coronal sections through the anteriormost orbit, the globe is cut through the level of the eyelids. In the midline of the orbital roof, below the frontal lobes of the brain, is the frontal sinus. The anterior segment of the globe may appear as several concentric densities that represent the cornea, lens, and anterior sclera. In the superior medial corner of the orbit are the trochlea and tendon of the superior oblique muscle. Inferiorly, the inferior oblique muscle can be seen as a linear shadow that runs from the inferomedial orbital wall toward the lateral orbit.
CORONAL SECTION THROUGH THE ANTERIOR ORBIT.
Sections cut through the anterior midorbit pass through the globe near its equator. At this level, the orbital roof is seen as a thin, curved plate of bone with an upper surface that undulates against the overlying frontal lobes. In the midline is the crista galli, and on either side are the cribriform plate and roof of the ethmoid sinus.
The orbital floor is a thin plate of bone that extends from the lowermost extent of the lamina papyracea and slopes downward and laterally to the inferolateral orbital wall. Immediately below the floor is the triangular maxillary sinus.
Centrally, the globe is seen to fill most of the orbital space ( Figs. 84-8 and 84-9 ). Superiorly, the thin superior rectus muscle lies adjacent to the globe, and above it is the levator muscle. Medially, the flattened medial rectus muscle lies within the orbital fat between the lamina papyracea and the globe. Just below the eye is the inferior rectus muscle, and laterally is the lateral rectus muscle. In the superomedial corner, a small, round shadow is the superior oblique muscle. At the medial edge of the superior rectus–levator muscle complex is a round enhancing structure, the superior ophthalmic vein.
CORONAL SECTION THROUGH THE CENTRAL ORBIT.
In coronal sections behind the globe, the orbital walls appear as in more anterior sections. Within the central space of the orbit lies the round optic nerve cut in cross section ( Figs. 84-10 and 84-11 ).
Figure 84-8 Coronal CT scan through the anterior portion of the orbit. The globes are cut in cross section, and the rectus muscles are seen as flattened densities near the sclera. IR, Inferior rectus; LM, levator muscle; LR, lateral rectus; MR, medial rectus; SR, superior rectus.
Figure 84-9 Coronal MRI scan through the anterior portion of the orbit. The cut is similar to that seen in Figure 84-8 . IR, Inferior rectus; LM, levator muscle; LR, lateral rectus; MR, medial rectus; SR, superior rectus.
On MRI scans, the central nerve can easily be distinguished from the nerve sheaths; the two are separated by the clear subarachnoid space. The four rectus muscles are seen against their respective orbital walls cut across their midbellies. The levator muscle appears as a separate thin strap just above and medial to the superior rectus. Above the medial rectus muscle, along the superomedial corner or the orbit, is the superior oblique muscle. The superior ophthalmic vein is a small enhancing circular density between the optic nerve and the superior rectus muscle, en route to the lateral orbit.
CORONAL SECTION THROUGH THE ANTERIOR ORBITAL APEX.
Toward the apex, the bony orbit narrows to a triangular section. Inferolaterally, the contour opens into the inferior orbital fissure that communicates with the infratemporal fossa. Within the orbit, the optic nerve, the superior oblique muscle, and all four rectus muscles can still be identified as separate structures. The superior ophthalmic vein is found more laterally, to the lateral edge of the superior rectus muscle. The ophthalmic artery is seen just above the optic nerve as it crosses over the nerve from lateral to medial.
CORONAL SECTION THROUGH THE POSTERIOR ORBITAL APEX.
At this level, the orbit is reduced to a small rounded space, open inferiorly to the pterygopalatine fossa. It is bounded laterally by the greater wing of the sphenoid and medially by the body of the sphenoid adjacent to the sphenoid sinus. Superolaterally, the orbit opens into the middle cranial fossa through the superior orbital fissure.
Figure 84-10 Coronal CT scan through the midorbit. The optic nerve lies centrally, surrounded by the rectus muscles. Just below the lateral edge of the superior rectus is the superior ophthalmic vein.
Figure 84-11 Coronal MRI scan through the midorbit. Structures are similar to those observed in Figure 84-10 .
Echography, or ultrasonography, is a technique that utilizes high-frequency sound waves to image tissue. Sound waves are generated by a piezoelectric crystal and are directed through the tissues of interest. Here they behave as light does—they demonstrate density-dependent velocity, as well as reflection, refraction, and scatter. Tissue characteristics such as cell compaction and reflective surfaces determine the specific echographic pattern.
Sound waves in the range of 8–10?mHz pass through orbital tissues from the probe tip. As the waves encounter tissues of different density and reflectivity, some are deflected and refracted, but others are reflected back to the probe, where they are detected and displayed on an oscilloscope. The amount of energy returned and detected by the probe determines the height (A-scan) or intensity (B-scan) of the resultant image. Thus, echography produces an image in which contrast reflects differences in the reflectivity of sound waves. The more reflective the interfaces within a tissue, the greater the amount of sound energy returned to the probe ( Fig. 84-12 ).
Figure 84-12 Physical characteristics of sound waves in diagnostic echography.
Figure 84-13 A-scan echographic image of the normal orbit. ES, Time scale in microseconds; IS, initial spike; O, orbital fat; R, retina; S, sclera; V, vitreous cavity.
Echography has the advantages of ease of use, no ionizing radiation, excellent tissue differentiation, and cost effectiveness. However, sound waves at 8–10?mHz do not penetrate beyond the midorbit. Also, echography cannot image both orbits simultaneously and is difficult for a nonspecialist to interpret.
The A-scan echographic image shows a one-dimensional array of spikes displayed along a baseline ( Fig. 84-13 ). The height of the spikes represents the signal strength or amplitude of the reflected echo, whereas the horizontal spacing between the spikes is dependent on the time required for the sound to reach its target and return to the receiver. This distance is proportional to the distance from the probe to the tissue target.
For orbital examinations, a standardized tissue sensitivity is used. Lesions are identified by their patterns, which are distinct from normal orbital echographic patterns. This technique is best for detecting tissue differentiation, surface characteristics, reflectivity, and vascularity.
On the normal orbital A-scan, the initial spike (see Fig. 84-13 , far left) represents echoes generated in the probe tip. The vitreous cavity is typically echolucent at baseline. The posterior pole
Figure 84-14 B-scan echographic image of the normal orbit in axial orientation. O, Orbital fat; ON, optic nerve; V, vitreous cavity.
complex is seen as a sharply rising perpendicular spike that represents the retina, followed by a slightly lower narrow zone (choroid) and a double-peaked maximally elevated spike that represents the inner and outer walls of the sclera. The orbital fat is seen as a descending chain of high irregular spikes, with rapid attenuation. Orbital lesions may appear as a zone of different reflectivity within the fat signals.
The B-scan echographic image shows a two-dimensional array that represents a “slice” through the tissue ( Fig. 84-14 ). The intensity or brightness of echoes is proportional to the strength of the sound waves returned to the detector. As with the A-scan, the spacing between echoes from left to right on the screen represents time or, more importantly, distance from the probe tip to the target tissue.
The basic B-scan examination requires scans in the transverse, longitudinal, and axial orientations, with visualization along all the orbital walls and the meridians of the globe. For orbital scans, the sound waves are usually directed through the eye. For some anteriorly placed lesions, a paraocular approach—with the sound beam passing beside, rather than through, the globe—may be more useful to demonstrate a lesion in relation to the eye. The B-scan is best used for the evaluation of topography, localization, and general contour.
The normal orbital B-scan image shows a cross section of the targeted area with a lucent (black) circular region that represents the globe (see Fig. 84-14 ). Behind this is a zone of brightly reflective echoes from the fat and interlobular septa. Several areas of lucency may represent the optic nerve or extraocular muscles.
1. Dutton JJ. Atlas of clinical and surgical orbital anatomy. Philadelphia: WB Saunders; 1994:201–40.
2. Sullivan JA, Harms SE. Characterization of orbital lesions by surface coil MR imaging. Radiographics. 1987;7:19–27.
3. Schenck JF, Hart HR, Foster TH, et al. Improved imaging of the orbit at 1.5T with surface coils. AJR Am J Roentgenol. 1985;144:1033–6.
4. Schenck JF, Hart HR, Foster TH, et al. High resolution magnetic resonance imaging using surface coils. In: Kressel HY, ed. Magnetic resonance annual 1986. New York: Raven; 1986:123–60.
5. Simon J, Szumowski J, Totterman S, et al. Fat-suppression MR imaging of the orbit. AJNR Am J Neuroradiol. 1988;9:261–8.
6. Dortzbach RK, Kronish JW, Gentry LR. Magnetic resonance imaging of the orbit. Part II. Clinical applications. Ophthal Plast Reconstr Surg. 1989;5:160–70.
7. DePotter P, Flanders AE, Shields CL, Shields JA. Magnetic resonance imaging of orbital tumors. Int Ophthalmol Clin. 1993;33:163–73.
8. DePotter P, Shields JA, Shields CL. MRI of the eye and orbit. Philadelphia: JB Lippincott; 1995:3–17.
9. Daniels DL, Pech P, Mark L, et al. Magnetic resonance imaging of the cavernous sinus. AJR Am J Roentgenol. 1985;145:1145–6.
10. Daniels DL, Yu S, Pech P, Haughton VM. Computed tomography and magnetic resonance imaging of the orbital apex. Radiol Clin North Am. 1987;25:803–17.
11. Hammerschlag SB, O’Reilly GVA, Naheedy MH. Computed tomography of the optic canals. AJNR Am J Neuroradiol. 1981;2:593–4.
12. Atlas SW, Grossman RI, Hackney HI, et al. STIR MR imaging of the orbit. AJR Am J Roentgenol. 1988;151:1025–30.
13. Atlas SW, Galetta SL. The orbit and visual system. In: Atlas SW, ed. Magnetic resonance imaging of the brain and spine. New York: Raven Press; 1991:709–22.
14. Harms SE. The orbit. In: Edelman RR, Hesselink JR, eds. Clinical magnetic resonance imaging. Philadelphia: WB Saunders; 1990:598–603.
15. Tien RD, Chu PK, Hesselink JR, Szumowski J. Intra- and paraorbital lesions: value of fat-suppression MR imaging with paramagnetic contrast enhancement. AJNR Am J Neuroradiol. 1991;12:245–53.
16. Dutton JJ, Klingele TG, Burde RM, Gado M. Evaluation of the suprasellar cistern by computed tomography. Ophthalmology. 1982;89:1220–5.
17. DeMarco JK, Bilaniuk LT. Magnetic resonance imaging: technical aspects. In: Newton TH, Bilaniuk LT, eds. Modern neuroradiology, vol 4, Radiology of the eye and orbit. New York: Raven Press; 1990:1–14.
18. Dortzbach RK, Kronish JW, Gentry LR. Magnetic resonance imaging of the orbit. Part I. Physical principles. Ophthal Plast Reconstr Surg. 1985;5:151–9.
19. Gore JC, Emery EW, Orr JS, Doyle FH. Medical nuclear magnetic resonance imaging: I. Physical principles. Invest Radiol. 1981;16:269–74.
20. Horowitz AL. MRI physics for radiologists, 2nd ed. New York: Springer-Verlag; 1992:4–54.
21. Sassani JW, Osbakken MD. Anatomic features of the eye disclosed with nuclear magnetic resonance imaging. Arch Ophthalmol. 1984;102:541–6.
22. Zonneveld FW, Koornneef L, Hillen B, de Slegte RGM. Normal direct multiplanar CT anatomy of the orbit with correlative anatomic cryosections. Radiol Clin North Am. 1987;25:381–407.
23. Dutton JJ. Radiographic evaluation of the orbit. In: Doxanas MT, Anderson RL, eds. Clinical orbital anatomy. Baltimore: Williams & Wilkins; 1984:35–56.
24. Beyer-Enke SA, Tiedemann K, Görich J, Gamroth A. Dünnschichtecomputertomographie der Schädelbasis. Radiologe. 1987;27:483–8.
25. Langer BG, Mafee MF, Pollock S, et al. MRI of the normal orbit and optic pathway. Radiol Clin North Am. 1987;25:429–46.
26. Artmann H, Grau A, Lösche CC. Aussagefähigkeit der Computertomograpgie der Ophthalmologischen Diagnostik. Radiol Diagn. 1989;30:621–8.
27. Byrne SF, Green RL. Ultrasound of the eye and orbit. St. Louis: Mosby–Year Book; 1992:1–51.
28. Byrne SF. Standardized echography. Part I: A-scan examination procedures. Int Ophthalmol Clin. 1979;19:267–75.