Magnetic Resonance Imaging of the Lacrimal Drainage System

Peter A. D. Rubin, MD, Jurij R. Bilyk, MD, John W. Shore, MD, F ACS, Francis C. Sutula, MD, Hong-Ming Cheng, PhD

Background: Magnetic resonance imaging (MRI) has excellent spatial and soft tis­sue resolution in the periocular and orbital region, especially when combined with surface coil techniques. Other methods, including computed tomography, dacryocystography, and dacryoscintigraphy provide limited information in the area of the lacrimal drainage system. Magnetic resonance imaging was used in conjunction with other imaging mo­dalities to compare the anatomic detail and clinically relevant information obtained about various pathologic processes in the lacrimal drainage area.

Methods: Patients with a variety of lacrimal drainage disorders underwent MRI with either head or surface coils. Contrast agents also were used in selected cases. Seven case reports are presented.

Results: Magnetic resonance imaging provided detailed information about local anatomy, extent of pathology, and, in some cases, etiology of the pathology.

Conclusions: Magnetic resonance imaging with surface coils provides detailed soft tissue information when compared with dacryocystography and computed tomography in the area of the lacrimal drainage system. In selected cases of lacrimal drainage system pathology, MRI may provide data that affect patient management. Ophthalmology 1994;101:235-243

Over the years, there have been numerous diagnostic techniques used to evaluate the lacrimal drainage system. These include probing, irrigation, dacryocystogram, and dacryoscintigraphy.l Each of these techniques relies on an indirect assessment of the lacrimal drainage apparatus. Intralacrimal endoscopy allows a direct view of the lac­rimal drainage system.2,3 However, the resolution is lim­ited, and in this setting it is not possible to evaluate the bony and soft tissue structures surrounding the lacrimal drainage system simultaneously. Computed tomographic (CT) scans of the lacrimal system offer improved reso­lution, especially of bony detail, yet the soft tissue contrast

Originally received: March 8, 1993. Revision accepted: July 28, 1993.

From the Department of Ophthalmology, Massachusetts Eye & Ear In­firmary, Harvard Medical School, Boston.

Presented as a poster at the American Academy of Ophthalmology An­nual Meeting, Dallas, November 1992.

Each author states that he has no proprietary interest in the development or marketing of the equipment or contrast agents named in this article.

Reprint requests to Peter A. D. Rubin, MD, Eye Plastics and Orbit Center, Massachusetts Eye & Ear Infirmary, 243 Charles St, Boston, MA 02114.

is limited.4 In addition, resolution is dependent on direct planar imaging.

Magnetic resonance imaging (MRI) offers several poten­tial advantages over other methods for evaluating the lac­rimal drainage system thilt may be useful in selected patients. Manipulation of the components of the pulse sequences, such as repetition time (TR), interpulse delay (Tl), echo time (TE), and tip angle optimize the differentiation of soft tissue signals and occasionally offer tissue-specific properties that help clarify the diagnosis.5 The relatively superficial 10-cation of the lacrimal drainage apparatus allows it to be well imaged with small surface coils.6 Improved spatial resolution is possible with a small field of view (up to but not limited to 0.3 X 0.3 X 3 mm) in conjunction with small surface coils which improve the signal-to-noise ratio and decrease the imaging time.7 Volumetric acquisition of signals with MRI permits the investigator to view the tissue in any plane of sectioning without loss of the image quality. Finally, MRI requires no exposure to ionizing radiation, unlike dacryo­cystography or CT. To determine the efficacy of MRI for the lacrimal drainage apparatus, we have applied high-res­olution MRI surface coil imaging in selected patients who had neoplastic, inflammatory, and obstructive disorders in this area.

235

Ophthalmology Volume 101, Number 2, February 1994

Methods

Using a General Electric Signa clinical imager (General Electric Medical Systems, Milwaukee, WI) with a field strength of 1.5 Tesla, MR.I scans were obtained. Patients were placed in the supine position. When a surface coil was used, a custom-made 1.5-inch (3.8 cm) diameter sur­face coil (Fig 1) was centered over the lacrimal sac on the side of interest. In case 4, a head coil was used in place of the surface coil. In case 7, a Siemens Magneton clinical imager (Siemens Medical, Iselin, NJ) with 1.5-Tesla field strength and a 5.0-inch (7.6 cm) diameter surface coil were used. In all cases, TRITE of 600/20 mseconds and a 256 X 256 acquisition matrix were used, with a 3- or 4-mm slice thickness. If T2-weighted images were ob­tained, TRITE of 2000/200 mseconds was used

When a contrast agent was applied, gadolinium-DTPA was administered intravenously as a bolus of 0.1 mmol/ kg or directly infused into the lacrimal sac after dilution of the contrast agent with sterile water until the final con­centration was approximately 300 mOsm/1.

Case Reports

Case 1: Intralacrimal Papilloma with Extension into the Nasolacrimal Duct. A 65-year-old man had a progres­sively growing lesion in the area of the right medial canthus of 3 years' duration (Figs 2A and 2B). The patient had undergone a previous excision of the lesion 4 years before presentation and was told that the growth "had something to do with the tear duct and was benign." On examination, a fullness in the area of the lacrimal sac was palpated. Results of nasal examination showed no abnormalities of the inferior meatus or turbinate. Dacryocystography could not be performed because of efface­ment of the puncta by the mass. Computed tomography dem­onstrated a mass in the area of the right lacrimal sac, with no evidence of bony erosion through the lacrimal sac fossa (Fig 2C). However, extension into the nasolacrimal duct could not be assessed. Incisional biopsy of the lesion was consistent with in­verted papilloma.

To better define the extent of the lesion for surgical planning, an MRI with surface coils was obtained (Figs 2D and 2E). On Tl-weighted imaging, a mass involving the entire lacrimal sac and extending through the nasolacrimal duct to a level just above the inferior meatus was noted. To maximally resect this lesion, which has a high propensity for recurrence and carcinomatous transformation,S a wide resection of the canaliculi, the lacrimal sac, and the nasolacrimal duct to the level of the inferior meatus was performed. All surgical margins proved negative on histo­pathologic review. In this case, MRI provided definitive evidence of nasolacrimal duct involvement, and provided a guide for complete resection.

Case 2: Intralacrimal Papilloma with Extension through the Punctum. A 38-year-old woman was noted to have a lesion at the left superior punctum of unknown duration. Clinically, the lesion appeared papillomatous (Fig 3A). To de­termine the intralacrimal origin and extent of the lesion, a dac­ryocystogram was obtained. The abnormal shadow-casting of the dye within the lacrimal sac suggested that there was extensive involvement of the lacrimal sac by a pathologic process (Fig 3B). Extirpative surgery was offered to the patient, but she refused.

236

Figure 1. The 1.5-inch surface coil used in this study.

Instead, she wanted to further investigate the extent of the lesion. Based on our experience with MRI with intralacrimal papilloma in case I, an MRI was obtained.

High-resolution T I-weighted axial surface coil imaging per­mitted localization of the papilloma within the canalicular sys­tem, and also demonstrated that the lesion did not extend deeply into the nasolacrimal duct (Fig 3C). The lesion appeared rela­tively hyperintense to the surrounding soft tissue. Additionally, after volumetric sampling in the region ofthe lacrimal drainage system, parasagittal sectioning in the plane of the nasolacrimal duct and sac was possible (Fig 3D). The hyperintense signal of the tumor on Tl-weighted imaging was atypical4 and may rep­resent mucinous accumulation or the presence ofmethemoglo­bin (blood-breakdown products) in the vicinity of the mass.9

Although surgery once again was recommended, the patient opted for a nonsurgical course.

Case 3: Distinguishing Anterior Ethmoidal Muco­pyocele from Lacrimal Sac Abscess. A l2-year-old girl had a 5-day history of swelling, rednesS, and pain in the region of the right lacrimal sac, without tearing or punctal discharge (Fig 4A). She had had similar symptoms 6 months earlier. The patient was started on antibiotics (ampicillin/clavulanate) for presumed dacryocystitis. After a favorable clinical response to the oral antibiotics, lacrimal probing and irrigation demonstrated a totally patent nasolacrimal drainage system. A dacryocysto­gram was obtained to assess for a possible diverticulum of the lacrimal sac or a dacryolith that could account for this patient's unusual presentation (Fig 4B). The dacryocystogram showed a normal contour and patency of the lacrimal drainage system. A CT scan demonstrated pansinusitis and a residual soft tissue mass in the region of the lacrimal sac (Fig 4C). A clearly de­monstrable lacrimal sac was not noted, nor was a dehiscence within the ethmoidal air cells surrounding the lacrimal sac ap­preciated. When the abscess recurred I month later, aspiration of its contents showed Hemophilus irifluenzae. a common sinus pathogen. An MRI was obtained to help determine if the origin of the abscess was from the lacrimal sac or the anterior ethmoid air cells. The Tl-weighted MRI scan showed that the soft tissue mass in the region of the lacrimal sac communicated directly with an opacified anterior ethmoid air cell (Fig 4D). In addition, there was evidence on the MRI scans that the lacrimal sac was not involved in the infectious process. The· patient's condition was addressed surgically by an external approach to a limited anterior ethmoidectomy. At surgery, a large dehiscence was noted within the lacrimal sac fossa. Presumably, this provided a path­way for anterior tracking of the chronically infected loculated anterior ethmoid air cell. The patient has had no evidence of a recurrent abscess and remains asymptomatic.

Rubin et al . MRI of the Lacrimal Drainage System

Figure 2. A, clinical photograph of the right lacrimal sac mass. B, on closer examination, the lesion encompasses the entire medial commisure. C, axial computed tomographic views at the level of the lacrimal sacs (top) and inferior nasolacrimal ducts (bottom). A lacrimal sac mass is evident (large arrow). However, asymmetry between the nasolacrimal ducts cannot be demonstrated (small arrows). D, an axial Tl-weighted magnetic resonance image at the level of the lacrimal sac clearly demonstrates the mass (arrows), relatively hypo intense to the orbital fat. E, an axial magnetic resonance image at the level of the nasolacrimal duct. The lesion is sti11 present (arrows).

Case 4: Chronic Lacrimal Sac Mucocele versus Lac­rimal Sac Neoplasm. A 70-year-old woman had tearing from the left eye and a painless mass in the region of the left medial

canthus for longer than 1 year. She denied any history of bloody tears or nasal discharge. The mass extended above the level of the medial canthal tendon and was nontender, noncompressible,

237

Ophthalmology Volume 101, Number 2, February 1994

Figure 3. A, a papillomatous lesion is seen protruding from the superior punctum (arrow). B, dacryocystogram showed abnormal contour of the lacrimal sac and possibly nasolacrimal duct. C, on magnetic resonance imaging, the lesion is seen extending into the lacrimal sac (arrow). More inferior images of the nascolacrimal duct showed no evidence of tumor extension. D, a parasagittal section in the plane of the lacrimal sac and superior nasolacrimal duct. The papilloma is apparent (arrows).

and nonerythematous (Fig 5A). Canalicular probing demon­strated a soft stop in the region of the common canaliculus, and irrigation resulted in total reflux from the punctum. The history and clinical findings raised the possibility of a lacrimal sac neo­plasm. Adjunctive studies were directed at determining if the mass was a neoplasm, in which case a dacryocystectomy without opening the lacrimal sac would be indicated, or a mucocele, with dacryocystorhinostomy as preferred surgical treatment. The CT scan showed a soft tissue mass in the lacrimal sac fossa with­out bony destruction (Fig 58).

An MRI with a head coil showed a homogenous mass with hyperintensity in both T I-weighted (Fig 5C) and T2-weighted (Fig 5E) images. The relative hyperintensity in the Tl-weighted image was consistent with mucus9

•10 (Fig 5C). After reformatting

the T I image through a fat-suppression algorithm, the lesion became hypointense along with the orbital fat (Fig 5D). After this MRI study, the lesion was believed to be a mucocele rather than a solid neoplasm. The patient underwent a dacryocysto-

238

rhinostomy. As expected, thick mucus, and not a solid mass, was encountered in the lacrimal sac at surgery.

Case 5: Gadolinium Infusion into the Lacrimal Sac. A 92-year-old woman had epiphora and a reducible mu­cocele of the lacrimal sac. Gadolinium, once diluted to 300 mOsm, was directly infused into the lacrimal sac through the lower canaliculus. On the T I-weighted image, gadolinium clearly outlines the lacrimal sac and demonstrates the anatomic rela­tionships between the lacrimal sac and the adjacent ethmoidal air cells (Fig 6). A fluid/gadolinium level is apparent within the fully distended lacrimal sac.

Case 6: Enlarged Nasolacrimal Duct after Silastic Intubation. A 30-year-old man underwent an uneventful re­pair of a canalicular laceration of the right lower eyelid with placement of a nasolacrimal stint (Crawford tubes, Jedmed In­strument Company, St. Louis, MO). After removal of the stint 4 months after the primary repair, the patient reported a whistling

Rubin et al . MRI of the Lacrimal Drainage System

Figure 4. A, the patient on initial presentation. B, the dacrocystogram shows normal contour of the lacrimal sac (solid arrow) and nasolacrimal duct (open arrow). C, axial computed tomographic scan shows opacification of the ethmoid air cells bilaterally, consistent with ethmoiditis. D, an axial magnetic resonance imaging demonstrates a soft tissue mass (large arrow) extending lateral to the lacrimal sac (small arrow). The mass directly communicates with an opacified anterior ethmoid air cell (open arrow).

sound emanating out of his puncta upon sneezing. On exami­nation, the lacrimal drainage system was patent. However, air was refluxed easily through the puncta when the patient exhaled against closed lips and pinched nostrils. Presumably, the delicate valves of the lacrimal system were disrupted upon the initial in­tubation and/or when the knot was rotated out of the nasolacrimal drainage system through the punctum as the silastic tube was removed. Although the valves of the lacrimal drainage system were not visualized directly, the MRI clearly demonstrated that the soft tissue of the lacrimal duct on the right side was more widely patent than the duct on the left side (Figs 7 A and 7B).

Case 7: Lacrimal Sac Mucocele versus N eo­plasm. An 85-year-old woman had a mass in the left medial canthal region, having noted enlargement over the previous 6 months (Fig 8A). She also complained of epiphora on the left side over the last 2 years, but denied blood-tinged tears or nasal discharge. On examination, a firm, rubbery mass was palpated just inferior to the medial canthal tendon in the area of the lacrimal sac. The lesion was not compressible, and no reflux through the puncta was noted. Results of nasal examination

were unremarkable. An axial CT scan obtained by the referring physician showed a cystic mass in the area of the left lacrimal sac with no evidence of bony erosion (Fig 8B). The nasolacrimal duct was not well visualized. An MRI with surface coils and fat suppression was obtained. T I-weighted images demon­strated a mass in the lacrimal sac without extension into the nasolacrimal duct (Fig 8C). The mass was relatively hyperin­tense, but not homogenous. Fat suppression caused decrease in the signal intensity of the lateral portion of the lesion (Fig 8D). On T2-weighted images, the mass remained hyperintense (Fig 8E). Based on the MRI scans, a mucocele was suspected. This diagnosis was confirmed during surgery when mucus was encountered upon opening the lacrimal sac. The mucus ap­peared less viscous than in case 4. Biopsy of the lacrimal sac was unremarkable.

Discussion

Both dacryocystography and dacryoscintigraphy provide shadow-casting of the lacrimal drainage system only,

239

Ophthalmology Volume 101, Number 2, February 1994

Figure 5. A, a mass in the area of the left medial canthus is apparent. B, axial CT shows a cystic mass without bony erosion (arrow). C, in Tl­weighted axial magnetic resonance images, the lesion is homogenously hyperintense, mirroring the signal of the adjacent orbital fat. D, the signal intensity of the mucocele drops off dramatically with fat suppression techniques. E, in T2-weighted views, the lesion again demonstrates hy­perintensity .

without providing any soft tissue detail. Computed to­mography, while excellent for bony detail, provides a more limited soft tissue detail than MRI, and suffers from image degradation in out-of-plane images. Magnetic resonance imaging offers several advantages for imaging of the na­solacrimal drainage system, including, among others, ex­cellent spatial resolution, direct images of soft tissues, multiplanar imaging without loss of resolution, and tissue­specific signal intensities.

240

The ability of MRI to image soft tissue directly as well as provide excellent spatial resolution with surface coil techniques allows exceptional soft tissue detail. In addi­tion, MRI provides better assessment of the extent of pathologic processes within the lacrimal drainage system than either CT11 or dacryocystography. In case 2, dacryo­cystography showed abnormal shadow-casting suggestive of a pathologic lesion within the lacrimal sac, but provided no further detail. Magnetic resonance imaging was more

Rubin et al . MRI of the Lacrimal Drainage System

Figure 6. The lacrimal sac is visible, with a gadolinium-fluid level dem­onstrated (arrows). At this concentration, the gadolinium has similar signal to fat on this Tl-weighted image.

anatomically specific. In case 1, the inferior extension of the papilloma was visualized above the ostium of the in­ferior meatus, and thereby allowed for appropriate surgical planning for complete tumor removal, while minimizing the removal of uninvolved tissue. In case 3, the visual­ization of the subcutaneous infectious process as a direct extension from the anterior ethmoidal cells rather than the lacrimal sac spared the patient unnecessary lacrimal drainage surgery.

The volumetric sampling inherent in the MRI tech­nique permits sectioning in any plane without loss of res­olution. In this fashion, the nasolacrimal duct which courses posterolaterally as it descends down toward the

inferior meatus can be imaged in its entirety through a single angulated parasagittal section. This advantage es­pecially is notable when compared with CT, in which out­of-plane reconstruction leads to considerable degradation of image resolution. (However, the emergence of new CT techniques, specifically spiral CT, may minimize this ar­tifact). In addition, the superficial location of the lacrimal drainage system allows the use of surface coil techniques, providing improved resolution over conventional MRI or head coil techniques.4

Finally, the ability ofMRI to differentiate the chemical content of lesions based on signal intensities in T 1- and T2-weighted images as well as fat-suppressed images also is helpful in certain circumstances. In cases 4 and 7, MRI was useful in distinguishing a lacrimal sac mucocele from a neoplasm, thereby altering the surgical management. The appearance of mucus on MRI is highly variable, de­pending on the viscosity and protein concentration of the mucus and relative amount of water. 10,12 As the water content of the mucocele decreases and the protein con­centration increases with time, the lesion becomes more hyperintense in T I-weighted images and less hyperintense in T2-weighted images. Thus, longstanding mucoceles tend to exhibit characteristic signal intensities which help to distinguish them from other lesions, as case 4 dem­onstrates. Unlike chronic, viscous mucus, the over­whelming majority of neoplasms, with the exception of melanotic melanoma and fat-containing tumors, have low signal intensity on T I-weighted imaging.4 It is also possible that, in cases 4 and 7, a neoplasm was present, and that bleeding in and around the tumor may have produced similar T 1- and T2-weighted images due to the presence of methemoglobin. 10 This alternative explanation is not likely, because neither patient complained of bloody tears or blood-tinged nasal discharge. On the other hand, mu­coceles of sl10rter duration are more difficult to distinguish from neoplasia, because the free water content is still high.1O This is well demonstrated in case 7, in whom the T I-weighted signal in the lacrimal sac was less homoge-

Figure 7. A. a Tl-weighted magnetic resonance image of the superior nasolacrimal duct. The right duct (solid arrow) is distended when compared with the asymptomatic left duct (open arrow). B. a section of the inferior portion of the nasolacrimal duct. The distended symptomatic side (solid arrow) again is compared with the contralateral duct (open arrow). The flow void of the angular arteries is an incidental finding (small arrows).

241

Ophthalmology Volume 101, Number 2, February 1994

Figure 8. A, clinical photograph demonstrates fullness in the area of the left lacrimal sac. B, an axial computed tomographic scan shows a cystic lacrimal sac mass without bony erosion. C, Tl-weighted magnetic res­onance axial images demonstrate a relatively hyperintense mass (arrow). D, with fat-suppression techniques, the hyperintensity decreased (arrow). E, T2-weighted image shows continued signal hyperintensity (arrow). Notice the resolution degradation from motion artifact in all three mag­netic resonance images.

nously distributed, and, although still suggestive of mu­cocele, the findings were less conclusive than in case 4. Interestingly, on gross examination, the mucus in case 7 appeared less viscous than in case 4, consistent with the intensity of the MRI signals. Despite certain trends in the appearance of mucus on MRI, the clinician must recog­nize that the signal intensity may be inconsistent, reflecting the highly variable composition of the mucus.

242

Case 7 also demonstrates one of the disadvantages in­herent in MRI. Any movement on the part of the patient results in motion artifact in the final image, seen as ghost­ing and degradation of resolution. Surface coils are es­pecially susceptible to motion-induced distortions. This effect is, in general, less of an issue with the shorter scan times required to obtain T I-weighted images but more noticeable with T2-weighted images.7 In addition, the

Rubin et al . MRI of the Lacrimal Drainage System

proximity of such dynamic structures as the eyelids and globe to the lacrimal sac may potentially increase motion artifact when imaging this area, especially if the patient blinks excessively or does not fixate on a specific target. Therefore, MRI may be less helpful in evaluating lacrimal drainage pathology in uncooperative patients. However, since both CT and dacryocytography require a degree of patient cooperation, motion artifact in MRI is a relative and variable disadvantage when compared with the other two modalities.

Magnetic resonance imaging also is less specific for im­aging bony structures than CT. Because of the adjacency of the ethmoid sinuses and nasal cavity to the lacrimal sac and nasolacrimal duct, accurate interpretation of bony detail with MRI is compromised further, since the thin bone may have the same signal intensity as the nearby aerated anatomy, especially in wide window settings. Thus, if a detailed view of the bony anatomy is essential for diagnosis and management, CT is a helpful comple­mentary study to MRI. In this series of patients, the signal from the mucosal lining of the ethmoid air cells and the nasal cavity provided adequate contrast and detail for as­sessing the extent of the pathologic processes (Fig 3E).

Shading artifact, a well-known effect of surface coils caused by rapid signal loss outside the coil radius,7 did not affect the image quality in this series, mainly because of the superficial location of the entire lacrimal drainage system. In addition, because of the relative closeness of both drainage systems to one another, the abnormal sys­tem could be compared with the normal contralateral side when necessary. The use of a head coil in case 4 dem­onstrates the increased field of view and depth obtainable with this technique. However, with larger surface coils, resolution of more superficial structures decreases. "Over­shining" of structures by the intense signal from fat in T I-weighted images also may be potentially problematic with surface coils (Figs 2D and 6) but is overcome by fat suppression (Fig 5D). Varying the window settings during final photography onto the film also helps to minimize this artifact.7

In cases 5 and 6, the clinical diagnoses were clear, and MRI was performed for purely investigational reasons. In case 5, the dilution of gadolinium was critical for obtaining high signal intensity within the lacrimal sac. Iffull-strength gadolinium is used, the T I-weighted signal is shortened excessively, resulting in a signal void within the lacrimal sac. In case 6, the clinical suspicion of damage to the lacrimal sac valvular system was confirmed by the high resolution available with surface coils.

Despite certain drawbacks inherent to the technology, MRI with surface coil technique provided specific infor­mation in a variety of lacrimal drainage system disorders and allowed for more accurate management decisions. Ad­ditionally, it proved more sensitive than either dacryocys­tography or cr in selected cases. Although MRI with surface coils currently is not recommended as the initial imaging modality for assessing the lacrimal drainage system or in all

cases oflacrimal drainage disorders, it should be considered as a useful adjunctive modality in assessing a selected subset oflacrimal drainage system pathology. The clinician should recognize that the vast majority of lacrimal drainage disor­ders may be diagnosed and successfully managed using a complete clinical examination, including probing, irrigation, and nasal examination, as well as conventional imaging techniques such as dacryocystography. Magnetic resonance imaging with surface coils is an emerging technology, and MRI in general is a relatively expensive diagnostic modality. It should, therefore, be reserved either as an investigational tool or for those patients who have difficult pathologic pro­cesses that defy diagnosis through more conventional tech­niques.

References

1. Dutton JJ. Diagnostic tests and imaging techniques. In: Linberg JV, ed. Lacrimal Surgery. New York: Churchill Livingstone, 1988;19-48.

2. Fein W, Daykhovsky L, Papaioannou T, et al. Endoscopy of the lacrimal outflow system. Arch Ophthalmol 1992; 110: 1748-50.

3. Singh AD, Singh A, Whitmore I, Taylor E. Endoscopic vis­ualisation of the human nasolacrimal system: an experi­mental study. Br J Ophthalmol 1992;76:663-7.

4. Bilaniuk LT, Schenck JF, Zimmerman RA, et al. Ocular and orbital lesions: surface coil MR imaging. Radiology 1985;156:669-74.

5. Wirtschafter JD, Berman EL, McDonald CS. Magnetic Resonance Imaging and Computed Tomography: Clinical Neuro-Orbital Anatomy. San Francisco: American Acad­emy of Ophthalmology, 1992. (Ophthalmology mono­graphs; 6)

6. Schenck JF, Hart HR Jr, Foster TH, et al. Improved MR imaging of the orbit at 1.5 T with surface coils. AJR Am J RoentgenoI1985;144:1033-6.

7. DeMarco JK, Bilaniuk LT. Magnetic resonance imaging: technical aspects. In: Newton TH, Bilaniuk LT, eds. Ra­diology of the Eye and Orbit. New York: Raven Press, 1990; chap 1.

8. Tanenbaum M, McCord CD Jr. The lacrimal drainage sys­tem. In: Tasman W, Jaeger EA, eds. Duane's Clinical Oph­thalmology, rev. ed. Philadelphia: JB Lippincott, 1992; vol. 4, chap. 13.

9. Van Tassel P, Lee YY, Jing BS, De Pena CA. Mucoceles of the paranasal sinuses: MR imaging with CT correlation. AJR Am J Roentgenol 1989;153:407-12.

10. Som PM, Dillon WP, Fullerton GD, et al. Chronically ob­structed sinonasal secretions: observations on T I and T2 shortening. Radiology 1989; 172:515-20.

II. Berkowitz RG, Grundfast KM, Fitz C. Nasal obstruction of the newborn revisited: clinical and subclinical manifes­tations of congenital nasolacrimal duct obstruction pre­senting as a nasal mass. Otolaryngol Head Neck Surg 1990; 103:468-71.

12. Bilaniuk LT, Zimmerman RA, Newton TH. Magnetic res­onance imaging: orbital pathology. In: Newton TH, Bilaniuk LT, eds. Radiology of the Eye and Orbit. New York: Raven Press, 1990; chap. 5; 5.48-5.55.

243