Search

Cholesteatoma is a progressive middle ear disease that can only be treated surgically but with a high recurrence rate. Depending on the extent of the disease, a surgical approach, such as microsurgery with a retroarticular incision or transcanal endoscopic surgery, is performed. However, the current examination cannot sufficiently predict the progression before surgery, and changes in approach may be made during the surgery. Large amounts of data are typically required to train deep neural network models; however, the prevalence of cholesteatomas is low (1-in-25, 000). Developing analysis methods that improve the accuracy with such a small number of samples is an important issue for medical artificial intelligence (AI) research. This paper presents an AI-based system to automatically detect mastoid extensions using CT. This retrospective study included 164 patients (80 with mastoid extension and 84 without mastoid extension) who underwent surgery. This study adopted a relatively lightweight neural network model called MobileNetV2 to learn and predict the CT images of 164 patients. The training was performed with eight divided groups for cross-validation and was performed 24 times with each of the eight groups to verify accuracy fluctuations caused by randomly augmented learning. An evaluation was performed by each of the 24 single-trained models, and 24 sets of ensemble predictions with 23 models for 100% original size images and 400% zoomed images. Fifteen otolaryngologists diagnosed the images and compared the results. The average accuracy of predicting 400% zoomed images using ensemble prediction model was 81.14% (sensitivity = 84.95%, specificity = 77.33%). The average accuracy of the otolaryngologists was 73.41% (sensitivity, 83.17%; specificity, 64.13%), which was not affected by their clinical experiences. Noteworthily, despite the small number of cases, we were able to create a highly accurate AI. These findings represent an important first step in the automatic diagnosis of the cholesteatoma extension. [ABSTRACT FROM AUTHOR]

Objectives: A broad mastoid extension limits cholesteatoma resection via a transmeatal approach including endoscopic ear surgery. Therefore, a preoperative diagnosis of mastoid extension is a the most critical factor to determine whether to perform mastoidectomy. The purpose of this study was to assess the efficacy of non-echoplanar diffusion-weighted imaging (non-EPI DWI) and T1-weighted imaging in the evaluation of mastoid extension in cholesteatomas of the middle ear.Methods: Patients who underwent magnetic resonance imaging (MRI) for pretreatment evaluation before primary surgery for pars flaccida or tensa cholesteatoma, which revealed a high-signal intensity in the mastoid on T2-weighed imaging were retrospectively evaluated. Two board-certified radiologists retrospectively evaluated the extent of cholesteatomas on MRI with non-EPI DWI, non-EPI DWI- and T1-weighted axial imaging. The presence of a high signal intensity on non-EPI DWI or low or high signal intensity on T1-weighted imaging in the mastoid was evaluated. All cases were subclassified as M+ (surgically mastoid extension-positive) or M- (surgically mastoid extension-negative).Results: A total of 59 patients with middle ear cholesteatoma were evaluated. There were 37 M+ cases and 22 M- cases. High-signal intensity on non-EPI DWI exhibited a sensitivity of 0.89 and specificity of 0.82, whereas partial low-signal intensity on T1-weighted imaging exhibited a sensitivity of 0.84 and specificity of 0.91 for detecting mastoid involvement. Complete high-signal intensity on T1-weighted imaging exhibited a sensitivity of 0.73 and specificity of 0.89 for detecting non-involvement of the mastoid. The sensitivity (0.92) and specificity (0.96) of combined non-EPI DWI and T1-weighted imaging evaluation were higher than those of with non-EPI DWI or T1-weighted imaging alone. The interobserver agreement for the presence of high-signal intensity in the mastoid cavity on non-EPI DWI was very good at 0.82, that of a partial low-signal intensity area in the mastoid cavity lesions on T1-weighted imaging was good, at 0.76 and that of complete high-signal intensity in the mastoid cavity lesions on T1-weighted imaging was good, at 0.67.Conclusions: The signal intensity on non-EPI DWI and T1-weighted imaging of the mastoid could be used to accurately assess the extent of middle ear cholesteatoma, which could facilitate surgical treatment planning. [ABSTRACT FROM AUTHOR]

Objectives: We aimed to clarify the usefulness of high-resolution computed tomography (HRCT) and establish HRCT criteria for presurgical assessment of the mastoid extension in pars flaccida cholesteatomas of the middle ear.Methods: Retrospective observational study. Patients who underwent primary surgery for pars flaccida cholesteatoma and those who underwent temporal bone HRCT for pretreatment evaluation were reviewed. The distance in the anterior-most portion of the mastoid sinus on HRCT was measured, and the presence of surgically verified mastoid extension of cholesteatoma was evaluated. All cases were subclassified as M+ (surgically mastoid extension-positive) or M- (surgically mastoid extension-negative).Results: A total of 107 patients with pars flaccida cholesteatoma were included. The distance in the M+ cases was significantly longer than that in the M- cases, and the cutoff value was 3.6 mm. The difference between the ipsilateral/diseased-side distance and the contralateral/evaluable side (difference value) in M+ cases was larger than that in M- cases, with a cutoff value of 0.6 mm. The inter-rater reliability of this distance measurement was excellent, regardless of imaging experience.Conclusions: The cutoff values of the distance and the difference value can be used for pretreatment HRCT evaluation of mastoid extension in middle ear cholesteatoma with relatively high accuracy, regardless of the experience and skill levels of the evaluator. [ABSTRACT FROM AUTHOR]

Background: The mastoid air cells are considered to develop better in children with congenital cholesteatoma but only a few quantitative studies have examined. Aims/objectives: The present study was undertaken in order to clarify the relationships between the development of the mastoid air cells in children with congenital cholesteatoma and clinical factors. Materials and methods: Mastoid air cell development was evaluated in 53 children with congenital cholesteatoma. The sizes of the mastoid air cells were measured on computed tomography scans of the temporal bone, and clinical factors, including age, the extent of invasion, and whether it was located close to the eustachian tube were evaluated. Results: The cross-sectional area of the mastoid air cells was significantly smaller on the affected side than on the unaffected side, especially in the older patients. The cases involving invasion into the mastoid portion also had smaller mastoid cells. Whether the cholesteatoma was located close to the eustachian tube had no effect. Conclusions and significance: The mastoid air cells on the affected side were less well developed. Our findings indicate that a high age and the presence of cholesteatoma in the mastoid region might be associated with suppressed mastoid pneumatization. [ABSTRACT FROM AUTHOR]

Background: The Japan Otology Society (JOS) proposed the classification and staging system for middle ear cholesteatoma. However, there was little analysis of the pathology of cholesteatoma using this staging system. Aims/objectives: To analyze the pathology of pars flaccida cholesteatoma using the staging system by JOS. Material and methods: A total of 183 cases of fresh pars flaccida cholesteatoma treated between January 2009 and December 2015 were included. We used the staging system recommended by JOS (2015). The association of the following variables in each stage of pars flaccida cholesteatoma was examined: age, gender, preoperative hearing level, staging, statuses of mastoid cell growth and stapes, tympanic sinus invasion. Results: Stage II disease showed the highest degree of progression, and peak incidence was observed in the third and fourth decades of life. Stage III significantly increased after the age of 40 years. The progression of the disease stage was significantly associated with deterioration of hearing level. Cholesteatoma invasion to tympanic sinus was recognized in 14.2% of cases. The state of the stapes is increasingly likely be to fracture as stage progresses. Conclusions: The JOS Staging System appropriately reflects the disease state, and it was found to be clinically meaningful in this study. [ABSTRACT FROM AUTHOR]

Background: Mastoid development, tympanic sinus depth, and residual disease after surgery for congenital cholesteatoma are probably related, but these relationships have not been examined in detail. Aims/objectives: This study aimed to clarify the relationships between the abovementioned factors. Materials and Methods: The subjects were 31 patients with congenital cholesteatoma (stage III or IV in Potsic's staging system) that underwent mastoidectomy. The cross-sectional area of the mastoid air cells was measured as described previously. Tympanic sinus depth was classified into A–C using Marchioni et al.'s system. Results: Patients with deep tympanic sinuses or residual disease exhibited significantly greater mastoid air cell development. However, little residual disease was found in the mastoid air cells. Conversely, residual disease was observed more frequently in the patients with deep tympanic sinuses. Conclusions and significance: After surgery for congenital cholesteatoma, residual disease is more likely to occur in patients with marked mastoid growth, possibly because they have deep tympanic sinuses. Cases in which congenital cholesteatoma spreads to the mastoid air cells are classified as stage IV in Potsic's system, but our findings indicate that invasion into a deep tympanic sinus is more important than invasion into the mastoid air cells. [ABSTRACT FROM AUTHOR]