Nerve Canals at the Fundus of the Internal Auditory Canal on High-Resolution Temporal Bone CT

PURPOSE: To identify and evaluate the normal anatomy of nerve canals in the fundus of the internal auditory canal which can be visualized on high-resolution temporal bone CT. MATERIALS AND METHODS: We retrospectively reviewed high-resolution (1 mm thickness and interval contiguous scan) temporal bone CT images of 253 ears in 150 patients who had not suffered trauma or undergone surgery. Those with a history of uncomplicated inflammatory disease were included, but those with symptoms of vertigo, sensorineural hearing loss, or facial nerve palsy were excluded. Three radiologists determined the detectability and location of canals for the labyrinthine segment of the facial, superior vestibular and cochlear nerve, and the saccular branch and posterior ampullary nerve of the inferior vestibular nerve. RESULTS: Five bony canals in the fundus of the internal auditory canal were identified as nerve canals. Four canals were identified on axial CT images in 100% of cases; the so-called singular canal was identified in only 68%. On coronal CT images, canals for the labyrinthine segment of the facial and superior vestibular nerve were seen in 100% of cases, but those for the cochlear nerve, the saccular branch of the inferior vestibular nerve, and the singular canal were seen in 90.1%, 87.4% and 78% of cases, respectively. In all detectable cases, the canal for the labyrinthine segment of the facial nerve was revealed as one which traversed anterolaterally, from the anterosuperior portion of the fundus of the internal auditory canal. The canal for the cochlear nerve was located just below that for the labyrinthine segment of the facial nerve, while that canal for the superior vestibular nerve was seen at the posterior aspect of these two canals. The canal for the saccular branch of the inferior vestibular nerve was located just below the canal for the superior vestibular nerve, and that for the posterior ampullary nerve, the so-called singular canal, ran laterally or posteolaterally from the posteroinferior aspect of the canal for the saccular branch. CONCLUSION: Five bony nerve canals in the fundus of the internal auditory canal were detected by high-frequency on high-resolution temporal bone CT. Familiarity with these structures can prevent confusion with, or misinterpretation as, a fracture line, and further study such as MR imaging may be required when any enlargement or erosion of these nerve canals is present.

Interactive Atlas Using Web Bowser: CT and MRI of the Temporal Bone

PURPOSE: The purposes of this study were to construct an interactive atlas of the temporal bone using a web-browser and to provide a template for web-based teaching files, using free and shared applets and scripts on the internet. MATERIALS AND METHODS: HRCT and MR images of the temporal bone including its normal anatomy, tumors, trauma, inflammation, anomalies and vascular diseases were used in this study. Acquired radiologic images were transformed to GIF/JPG formats and to achieve appropriate image quality, were retouched. Text and image files of normal anatomy and diseases were written by HTML. JavaScript and applets were inserted in the HTML files for the interactive display of images and texts. In order to review anatomic features and diseases, a sarch index was also attached to the last part of the file. RESULTS: Using interactive images and texts, temporal bone anatomy and disorders were displayed. Scripts and applets were also useful for indicating specific points of interest when a mouse was placed over the anatomic sites. The atlas may be viewed in the form of a CD-ROM, or via the internet using any computer platform or web-browser. CONCLUSION: This web-based teaching file of the temporal bone offers dynamic and interactive education. It can be usefully employed as a template for the production of interactive educational materials, offering JavaScript and Providing suitable input for classes. It can replace texts and imaging contents.

Absent Bow-Tie Sign in Knee MRI

PURPOSE: 'Absent bow-tie sign'is interpreted as positive when a bow-tie-shaped body segment is seen on only one or no slice of 4- or 5-mm thick sagittal images, and is a well known as a useful sign in diagnosing bucket-handle meniscal tears. In practice, however, we have found that this sign was also positive in certain cases other than bucket-handle tears. We have assumed that if the normal range of meniscal body width, as determined among Westerners, is transferred to the Korean population without verification and modification this might lead to misdiagnosis. The purpose of this study, therefore, is to examine the reliability of the 'absent bow-tie sign'. MATERIALS AND METHODS: Among 454 cases in which knee MRI had been performed, we retrospectively evaluated 862 menisci, the total remaining after cases of discoid meniscus or those involving previous meniscectomy had been excluded. Among the 862 menisci, 614 were normal, 97 showed degeneration, 43 showed buck-et-handle tearing, and 108 showed tears other than bucket-handle tear. In all cases, proton-denwity and T2-weighted images were obtained in both sagittal and coronal planes, with 3mm section thickness and 1mm gap. We recorded the number of sagittal images in which the body segment of each meniscus had a bow-tie appear-ance, and measured the width of each meniscal body, as seen on midcoronal images. RESULTS: In all cases but one of bucket-handle tears (97.7%), the bow-tie sign was absent, as it was in 73.2% ofnon-bucket-handle tears, 35.0% of degenerated menisci and 27.5% of normal menisci. Among the non-tear group, 56.4% of menisci in the female group and 27.1% in the male group had bodies less than 9mm wide. CONCLUSION: In the diagnosis of bucket-handle tears, the 'absent bow-tie sign'is a very sensitive indicator. It is nonspecific, however, and merely suggests some significant deficiency in the meniscus body or small menis-ci,so can be positive in other cases. Thus the interpreter should be aware of the characteristics of this sign especially when used to interpret MRI of the knee of a female Korean patient.

MRI of Ganglion Cysts in Uncommon Sites or with Atypical Appearance

Ganglion cysts are common lesions which are most often found around the hands and feet. Magnetic resonance (MR) imaging is seldom performed; cysts are identified incidentally on MR images obtained for the purpose of evaluating other disorders. Clinically complicated ganglions are often confused with other more serious tumors, and MR examination is thus requested. Ganglions arising from uncommon sites or showing unusual MR findings can lead to wrong diagnosis. Thus, a thorough knowledge of potential sites and of the MR findings of unusual ganglion cysts can prevent mistake s. We therefore describe the MR findings of unusual ganglion cysts, including complicated ganglions, cruciate ligament ganglion cysts, meniscal cysts, labral cysts, ganglions in shoulder notches and tarsal tunnels, and intermuscular ganglion cysts.

Bilateral Contributions of the Cerebellum to the Complex Motor Tasks on EPI fMRI

PURPOSE: To demonstrate activation signals within the cerebellar cortex and to determine the side of thecerebellar cortex eliciting activation signals in response to complex motor tasks, as seen on EPI fMRI MATERIALS AND METHODS: Seven right-handed subjects (M:F=3:4;mean age, 30.3 years) underwent repetitive finger appositionwith the dominant right hand. Using a 1.5T MRI scanner, EPI fMR images were obtained. MR parameters used for EPIfMRI were TR/TE/Flip angle : 0.96 msec/64 msec/90 degree, FOV 22cm, 128 X 128 matrix, 10 slices, 10mm thickness whilethose for SE T1 weighted localized images were TR/TE : 450/16, FOV 23cm, 256 X 256 matrix. The paradigm was threesets of alternate resting and moving fingers for six cycles, resulting in times of 360 seconds (10 slices X 15 EPIX 6 cycles = 900 images). Image processing involoved the use of a 200mHz Dual Pentium PC with homemade software.T-testing (p<0.005) and time series analysis were performed, and to verify the locations of activated regions,resulting images were analyzed in a color-coded overlay to reference T1-weighted spin echo coronal images.percentage change in signal intensity (PCSI) was calculated from the processed data. RESULTS: All normal subjectsshowed significant activation signals in both the contralateral (left) primary motor cortex (PCSI=3.12%+/-0.96) andipsilateral (right) cerebellar cortex (PCSI=3.09%+/-1.14). Signal activation was detected in the contralateralsupplemental motor area (2.91%+/-0.82), and motor activation in the anterior upper half of the contralateralcerebellum (PCSI=2.91%+/-0.69). The difference in activation signals between both sides of the cerebellar cortexwas not statistically significant. All data were matched with time-series analysis. CONCLUSION: Bilateralcerebellar activation is associated with unilateral complex finger movements, as seen on fMRI. This result maysupport the recent neurological observation that the cerebellum may exert bilateral effects on motor performance.

MRI of Bucket-Handle Tears of the Meniscus of the Knee

PURPOSE: To determine the frequency of the MRI signs of meniscal bucket - handle tears already known as the double PCL sign, the flipped meniscus sign, the absent bow-tie sign, and the fragment-in-notch sign, and to compare the sagittal with the coronal images. MATERIALS AND METHODS: We retrospectively rev i ewed the MR findings of 37 patients in whom an initial interpretation of MR images had suggested meniscal bucke t - handle tears. All underwent subsequent arthroscopic evaluation and in 28, bucke t - handle tears were confirmed. Sagittal double-echo and coronal fat-suppressed double-echo T 2 - weighted images were obtained. Sagittal images were evaluated to determine whether or not signs of bucket-handle tear were evident, and coronal images we r e checked for a torn meniscus with displaced fragment. We also evaluated the MR findings of the nine false positive cases. RESULTS: The prevalence rate of absent bow-tie, double PCL, fragment-in-notch, and flipped meniscus signs was 96.4%, 53.6%, 17.9%, and 10.7%, respective l y. The detection rate for displaced fragment was higher with coronal images (92.9%) than with sagittal images (78.6%). Among the nine false positive cases, a longitudinal tear in the discoid meniscus was most common. A false-positive diagnosis was much more frequent on sagittal than on coronal images. CONCLUSION: The prevalence rate of absent bow-tie sign was very high, but was accompanied by a relatively high rate of misinterpretation. Coronal fat-suppressed T2- weighted images provided more reliable clues for the diagnosis of bucket-handle tears, with a high detection rate of displaced fragment.

MRI of Bucket-Handle Tears of the Meniscus of the Knee

PURPOSE: To determine the frequency of the MRI signs of meniscal bucket - handle tears already known as the double PCL sign, the flipped meniscus sign, the absent bow-tie sign, and the fragment-in-notch sign, and to compare the sagittal with the coronal images. MATERIALS AND METHODS: We retrospectively rev i ewed the MR findings of 37 patients in whom an initial interpretation of MR images had suggested meniscal bucke t - handle tears. All underwent subsequent arthroscopic evaluation and in 28, bucke t - handle tears were confirmed. Sagittal double-echo and coronal fat-suppressed double-echo T 2 - weighted images were obtained. Sagittal images were evaluated to determine whether or not signs of bucket-handle tear were evident, and coronal images we r e checked for a torn meniscus with displaced fragment. We also evaluated the MR findings of the nine false positive cases. RESULTS: The prevalence rate of absent bow-tie, double PCL, fragment-in-notch, and flipped meniscus signs was 96.4%, 53.6%, 17.9%, and 10.7%, respective l y. The detection rate for displaced fragment was higher with coronal images (92.9%) than with sagittal images (78.6%). Among the nine false positive cases, a longitudinal tear in the discoid meniscus was most common. A false-positive diagnosis was much more frequent on sagittal than on coronal images. CONCLUSION: The prevalence rate of absent bow-tie sign was very high, but was accompanied by a relatively high rate of misinterpretation. Coronal fat-suppressed T2- weighted images provided more reliable clues for the diagnosis of bucket-handle tears, with a high detection rate of displaced fragment.

MR Findings of Nasal Cavity Lesions Showing the Infundibular Widening on CT

PURPOSE: Several kinds of nasal cavity lesions located in the region of the infundibulum on CT. At such time,the visualization of these lesions is very sinilar. The purpose of this study was to differentially diagnose thesenasal cavity lesions through evaluation of the MR findings. MATERIALS AND METHODS: In 51 cases of pathologicallyproven nasal cavity masses which on CT showed infundibular widening, we retrospectively evaluated the MR findings.The cases involved prolapsed antral mucosa from sinusitis(n=15), inverted papilloma(n=10), antrochoanalpolyp(n=10), aspergillosis(n=9), and nasal polyp(n=7). All patients underwent both CT and MR. imaging. RESULTS: In all cases, CT findings were similar ; soft tissue masses filling the maxillary sinus and nasal cavity wereassociated with infundibular widening caused by pressure on the uncinate process, leading to erosion. Differentialdiagnosis by CT was very difficult ; MR T2 weighted imaging was most effective for differential diagnosis of thesenasal cavity masses. Prolapsed antral mucosa showed central inhomogeneous mixed signal intensity, with aperipheral rim of hyperintensity along the sinus wall and nasal component. Antrochoanal polyps showed homogeneousbright signal intensity of the antral and nasal component. Aspergillosis showed central dark signal foci. Invertedpapillomas showed mixed intermediate and high intensity mixed with high signal intensity. Nasal polyps showedstriation mixed of intermediate and high signal intensity, while nasal polyp showed striation of intermediate andhigh intensity. On Gd-enhanced T1 weighted images, prolapsed antral mucosa and antrochoanal polyp showedperipheral rim enhancement of the antral and nasal component. In contrast, inverted papilloma and nasal polypshowed intense enhancement of the mass and can be separate from the sinus inflammatory disease. CONCLUSION: Various nasal cavity masses showing infundibular widening on CT can be differentiated on MR images, especially ofthese are T2 weighted or contrast enhanced T1 weighted.

CT and MR evaluation of Choanal Polyps

PURPOSE: To evaluate the various types and CT and MR imaging findings of choanal polyps. MATERIALS AND METHODS: We classified 42 cases of surgically proven choanal polyps into three types according to site of origin,and retrospectively analyzed imaging findings; in the case of antrochoanal polyps, particulr attention was givento extension of the lesion into the nasal cavity. We also determined whether MR imaging can provide moreinformation than CT. RESULTS: The antrochoanal polyp was the most common type( 33/42 : 78.6% ). Three cases ofchoanal polyps(7.1%) originated in the sphenoid sinus (sphenochoanal polyps), and while six( 14.3% ) did notinvolve the sinus( pure choanal polyps ). Thirty-six of 42 choanal polyps(85.7%) extended into either the choanaor nasopharynx. Three cases of antrochoanal polyps protruded through the middle meatus, and two into theoropharynx. The maxillary sinus component of an antrochoanal polyps extended more frequently through the accessoryostium( 29/33 ) than through the natural ostium. For delineating the stalk of an antrochoanal polyp, MR imagingwas not superior to CT. CONCLUSION: The most common type of choanal polyp is the antrochoanal, followed by purechoanal and sphenochoanal. CT and MR imaging can help identify lesion continuity between a choanal polyp and theparanasal sinus.

Verification of the Reliability of Sagittal MR Images in Assessment of Segmental Instability of the Wrist

PURPOSE: To determine the reliability of carpal angle measurement using sagittal wrist MR images for thediagnosis of dorsal intercalated segmental instability (DISI) or ventral intercalated segmental instability(VISI). MATERIALS AND METHODS: In ten asymptomatic adult volunteers, capitolunate, scapholunate, and radiolunate angleswere measured on lateral radiographs and five different series of sagittal MR images, and compared andstatistically analysed. The four MR series, obained in a prone position, with the wrists over the head, were asfollows : A, sagittal images parallel to the third metacarpal bone axis, in a neutral position ; B, obliquesagittal images parallel to the fourthe metacarpal axis, in a neutral position, ; C and D, sagittal imagesparallel to the third metacarpal bone axis, with the wrist at 15degree radially and ulnarly deviated positions,respectively. The fifth true sagittal MR image (E) was obtained in a supine position with the wrists locatedbeside the trunk, in a neutral position. RESULTS: In the MR series performed in the prone position, thecapitolunate and radiolunate angles measured in the series A, B and D were significantly larger than those seen onlateral radiographs. The scapholunate angle showed a statistically significant difference(p=.014) in series D,where one case showed DISI configuration at a scapholunate angle of 81degree. In other series, the angles were notsignificantly different from those seen on lateral radiographs. CONCLUSION: The lunate appears more dorsallytilted on sagittal MR images obtained in the prone position with the wrists above the head than on lateralradiographs, and a DISI configuration can be minicked. IF diagnosing wrist segmental instability by interpretingsagittal MR images without lateral radiographs, there is potential risk of misdiagnosis. It is believed that wristpositioning is crucial for the correct measurement of angles.

Cholesteatoma of the Temporal Bone: MR Findings

PURPOSE: The purpose of this study is to evaluate the diagnostic capability of MR imaging for distinguishingcholesteatoma from other causes of inflammation. MATERIALS AND METHODS: We prospectively evaluated the MR imagesof 17 patients with a fully opacified middle ear in whom CT of the temporal bone suggested cholesteatoma. Allpatients underwent middle ear surgery and cholesteatoma was verified in 15 cases, granulation tissue in 15,cholesterol granuloma in four and hemotympanium in one. For each lesion, MR and pathologic findings werecorrelated. RESULTS: All cases of cholesteatoma showed low to intermediate signal intensity on T1-weightedimages. On T2-weighted images, 11 cases showed high signal intensity, while seven were less than CSF and four werethe same. Four cases revealed central low signal intensity with peripheral rim of high signal intensity. On GdDTPA-enhanced images, peripheral or marginal enhancement due to surrounding granulation tissue was noted in 12cases. Two appeared to be totally enhanced. One case of congenital cholesteatoma showed no enhancement. OnT2-weighted images, all cases of granulation tissue showed high signal intensity, and on neither T1- norT2-weighted images could these be distinguished from cholesteatoma. On Gd DTPA-enhanced study, all cases ofgranulation tissue were enhanced, and on T1WI, all cases of cholesterol granuloma showed homogeneous bright signalintensity on T1WI. One case of cholesterol granuloma in the mastoid showed bright signal intensity mixed withirregular areas of dark signal intensity. A further comparison of MR imaging with CT is that two case oflabyrinthitis and one of meningitis were diagnosed on Gd- enhanced T1-weighted images. CONCLUSION: Todifferentiate cholesteatoma from granulation tissue and cholesterol granuloma by the evaluation of their signalintensities and enhancement patterns, MR imaging is superior to CT. When there are cholesteatoma-associatedcomplications such as labyrinthitis or meningitis, postcontrast MR imaging can provide better information.

Magnetic Resonance Neurography with Short Tau Inversion Recovery Sequences for Cervical Radiculopathy

In magnetic resonance(MR) imagings with short tau inversion recovery(STIR) sequences fat signals are specifically suppressed. These imaging techniques, therefore, have been used to identify optic nerve lesions in the orbits, because there are a lot of fat tissues which may obscure the optic nerves in the orbit. The cervical spinal nerves or ventral primary rami of brachial plexus, during their courses in the necks between anterior and middle scalene muscles, are also surrounded by considerable amount of fat deposits. So we used these MR imaging techniques with already used posterior neck coil system, to image cervical spinal nerves directly, and to know whether there are certain signal changes in those nerves of the involved nerve roots, in 12 patients who were considered to have cervical radiculopathies by conventional diagnostic means. In MR neurography with STIR sequences(STIR MRN) of axial and coronal images, signals of the cervical spinal nerves of the involved roots were detected as fairly bright, and were discerned from signals of the spinal nerves of the uninvolved roots. We also measured contrast-to-noise ratio(CNR) of the spinal nerve signals of the involved and uninvolved roots in axial images. The average CNR value of the former was 4.48+/-0.889 and of the latter 1.40+/-0.274. Our fingings indicate that STIR MRNs can be helpful in limiting the lesions more accurately and in making surgical decisions for cervical radiculopathies caused by multileveled discs or spondylosis.

Comparison of MR Myelography with Conventional Myelography in Lumbar Disc Diseases

PURPOSE: To compare MR myelography(MRM) using heavily T2 weighted fast spin echo(FSE) with conventional myelography in the evaluation of lumbar degenerative disc diseases. MATERIALS AND METHODS: A total of 50 patientswith lumbar degenerative disc diseases underwent both MRM and conventional myelography. A 1.5 T Signa MR systemwas used. For MRM, heavily T2 weighted FSE pulse sequences were performed in sagittal, coronal and both obliqueplanes. The images of MRM and conventional myelography were analyzed retrospectively and compared, focusing on the pattern of compression of the thecal sac and nerve roots. RESULTS: MRM yielded comparable images to conventional myelography in 68% (34/50). MRM was superior to conventional myelography in 11 cases (22%), and the opposite resulted in 5 cases(10%). MRM could visualize the spinal canal below the level of complete block in 6 patients,provided better contrast in the cases of dural ectasia in 3 cases and better delineation of individual nerve roots with source images in 1 case. Additional informations such as incidental retroperitoneal vascular malformation and extravasation of the CSF could be obtained by MRM in 1 case each. Disadvantages of MRM were poor contrast in thecases of multiple spinal stenosis, obscuration of the nerve roots by engorged paraspinal vessels, and occassional difficulty in defining the levels due to vanishment of the bony background image. CONCLUSION: MRM is a safe, noninvasive, nonionizing modality which in very effective in evaluating the lumbosacral disc disease. Use ofFSE-MRM in addition to the routine MR imaging could be very helpful not only for better evaluation of nerve roots but also for the additional informations.

MR Imaging of the Knee: Three-Dimensional Fourier Transform GRASS Technique

PURPOSE: To evaluate the usefulness of three-dimensional(3D) Fourier transform(FT) gradient refocused acquisition in steady state(GRASS) technique for MR imaging of the knee. MATERIALS AND METHODS : Sixty-three kneesin 61 patients were imaged on the 1.5T MR system. We compared 3DFT GRASS technique with 2D spin echo(SE) techniquein terms of conspicuousness of the lesions of internal knee structures based on the results of arthroscopy or open surgery. As a SE technique, sagittal T1-and T2-weighted, and coronal fat-suppressed T2-weighted sequences were performed. Sixty contiguous axial scans with 0.7 or 1mm section thickness were performed using 3D GRASS technique, and we also evaluated arbitrarily reformatted images produced from the original axial voxel images. RESULTS: For the depiction of the tear, 3DFT GRASS was superior to 2D SE in three cases of medial meniscus, one of later almeniscus, and two of anterior cruciate ligament. Specificity of 3D GRASS was also higher than that af 2D SE inevaluation of lateral meniscus and anterior cruiciate ligament. There was no significant difference in MR diagnosis for tears of the posterior cruciate, medial collateral, and lateral collateral ligaments. 3D GRASS was superior in evaluating the extent and morphology of the torn menisci. CONCLUSION: The 3DFT GRASS technique was comparable or even superior to the 2D SE technique in the evaluation of the internal structure of the knee, andcan be expected to supplement standard MR knee techniques, especially in complicated cases of meniscal orligamentous tears.

Cochlear Otosclerosis (Otospongiosis): A Case Report

Otosclerosis occurs when the normal ivory-like endochondral bone of the otic capsule is replaced by foci of spongy vascular irregular new bone. Eventually these foci of haversian bone become more solid. There anr two typesof otosclerosis : fenestral and retrofenestral or cochlear. The cochlear type is much less common and rarely occurs without fenestral involvement. Otosclerosis occurs very rarely in blacks and orientals ; in otolaryngologic articles in Korea there are few reports of otosclerosis, and then only the fenestral type. This is a case reportof cochlear otosclerosis that can be diagnosed on temporal bone CT.

CT Evaluation of Maxillary Sinus Aspergillosis: Morphological Patterns on CT

PURPOSE: In order to classify the involving patterns of maxillary sinus aspergillosis on CT that will be an important guidance for functional endoscopic sinus surgery(FESS) and to evaluate CT findings of each pattern. MATERIALS AND METHODS: We reviewed CT findings of 37 cases of maxillary sinus aspergillosis which were confirmed pathologically. These were classified into three patterns according to the location:the maxillary pattern(I), the infundibular pattern(11), and combined pattern(Ill). All cases were correlated with FESS findings. We also evaluated CT density, presence of contrast enahncement and calcifications, and surrounding bony changes of each pattern. RESULT: The combined pattern was most commonly seen in 24 of 37 cases(64.9%). The maxillary pattern was identified in 8(21.6%), with 5(13.5%)cases demonstrating the infundibular patttern. These patterns were consistent with FESS findings. Calcifications were seen in 28 cases(75.7%). 31 cases(83.8%) revealed hypodensity and 16cases among them showed peripheral enhancement. Central high density with peripheral low density rim was shown in 25%(2/8) of the maxillary pattern and 12.5%(3/24) of combined patttern. Surrounding bony changes were noted mainly in infundibular pattern and combined pattern. Erosion of uncinate process showed all cases of both patterns. Erosion or elevation of ethmomaxillary plate was seen in 3 of 5 cases of infundibular pattern and all cases of the combined pattern. CONCLUSION: CT fidnings of maxillary sinus aspergillosis can be classified into 3 patterns according to their distribution. These classification could be helpful for planning FESS approach. Calcifications are the most characteristic findings of maxillary sinus aspergllosis whereas CT density or enhancement pattern do not contribute to diagnosis.

CT Findings of Paranasal Sinus Osteoma

PURPOSE: To evaluate the incidence and the site of osteomas and their relation to sinonasal inflammation. MATERIALS AND METHODS: 1578 PNS computed tomography (CT) of patients with symptoms of sinusitis were evaluated and correlation with plain radiographic finding was done. RESULTS: The total incidence of osteomas was 0.82%(13/1578) on plain radiography and 3.29% (52/1578) on CT. Fifty-seven osteomas were found in 1578 PNS CT, including 2 osteomas in 5 cases. The most common site was the ethmoid sinus(54%, 31/57) and the next was the frontal sinus(35%, 20/57) on CT. Forty-five osteomas (78.9%) were associated with the sinonasal inflammatory change. CONCLUSION: The total incidence of osteomas from our results was higher than previous reports. And based on the analysis of CT, the most common site was not the frontal sinus as all the previous reports stated, but the ethmoid sinus, probably by virtue of high resolution of CT and/or higher prevalence of inflammation in the ethmold sinus.

CT Evaluation of Nasal Cavity Masses: Differential Diagnosis between Nasal Polyps and Their Mimics

PURPOSE: The purposes of this article are to identify CT findings distinguishing nasal polyps, that are the most common cause of the nasal polypiod lesions, from other nasal cavity masses and to identify differential points between benign and malignant masses of nasal cavity. MATERIALS AND METHODS: We classified 567 cases of pathologically proved nasal cavity masses into 4 different groups on CT according to the sites of origin and surrounding bone changes. RESULT: The nasal polyps were the most common cause of nasal cavity masses(515/567). Group I had a high diagnostic specificity for the nasal polyps (513/515, 99.6%) although inverted papilloma (8/522) and malignant melanoma(1/522) showed similar CT appearances. The representitive nasal mass of Group Ila was the inverted papilloma (15/24, 66.5% of the inverted papilloma). Group III pattern was seen in 3 cases of benign minor salivary gland tumor and 2 cases of pyogenic granuloma. Group IV suggested malignancy and thus was mostly observed in malignant tumors except I case of inverted papilloma and 1 case of granulomatous necrosis. CONCLUSION: The differential diagnosis between nasal polyps and the other nasal cavity masses is possible by characteristic bone changes and their sites of origin revealed on CT. In addition, it is possible to differentiate malignant masses from benign in most cases.

Medial Depression with Bony Dehiscence of Lamina Papyracea as an Anatomic Variation: CT Evaluation

PURPOSE: To evaluate the incidence and CT findings of the medial depression and bony dehiscence of lamina papyracea as an anatomic variation. MATERIAL AND METHODS: 1472 PNS CTs of the patients with symptoms of chronic sinusitis were retrospectively evaluated. RESULTS: The total incidence of depressed lamina papyracea as an anatomic variation was 3.5%(52/1472) on PNS CT. There was a statistically significant correlation between the increasing age and the incidence of delamina papyracea. Depression of lamina papyracea anterior to the basal lamella were more common those of the posterior depression. Associated findings were herniation of adjacent fatty tissue in all cases and the roedial bowing and hypertrophied configuration of the medial rectus muscle without significant herniation in 19 cases(34%). CONCLUSION: Nontraumatic, asymptomatic depression with bony dehiscence of lamina papyracea as an anatomic variation is not uncommon with the incidence of 3.5%. Recognition of its existence and degree may helpful in avoiding various ocular complication during ethmoid surgery.

Anatomic Variations on PNS CT

PURPOSE: To illustrate frequently encountered anatomic variations on PNS computed tomography(CT) and to assess their clinical significance. MATERIALS AND METHODS: CT findings of 1523 patients, who underwent PNS CT with no history of sinus surgery, were reviewed, and prevalence of each anatomic variation and its relationship with mucociliary clearance pathway was evaluated. RESULTS: Two categories of anatomic variations were encountered. The first group included variations of the turbinates, such as concha bullosa(28.6% ), paradoxical middle turbinate(31.5%), of the uncinate process, such as medially bent or folded(36.3%), laterally bent(3.7%) or uncinate bulla(0.5%), of the ethmoidal sinus, such as Haller's ce11(28.2%), large agger nasi ce11(9.6%) or large ethmoidal bulla(23.5%) and nasal septal deviation(24. 1%), which might cause obstruction of mucociliary clearance pathway and thus give rise to secondary obstructive sinusitis. The second group included Onodi ce11(1.4%) and medial depression of the lamina papyracea (3.5%), which were not related with obstructive sinusitis. 87 cases of large agger nasi cell were associated with obstruction of nasofrontal recess. Large ethmoidal bulla(452 cases), Haller's ce11(245 cases) and true concha bullosa (25 cases) contributed to narrowing of the infundibulum. Medially bent uncinate process(220 cases), concha bullosa(157 cases), paradoxical middle turbinate(126 cases) and nasal septal deviation(93 cases) were observed in middle meatal obstruction and supreme concha (3 cases) accompanied sphenoethmoidal recess obstruction. CONCLUSION: Recognition of anatomic variations on PNS CT is important to build a treatment plan in patients with obstructive sinusitis as they may represent causes of the disease, and to avoid critical complications during functional endoscopic sinus surgery.