Abstract

Objective

Non-echo planar diffusion-weighted imaging sequence and fusion imaging are increasingly used in the diagnosis of cholesteatoma. But it is still challenging to locate cholesteatomas and differentiate from other opacifications. This study aimed to evaluate whether the exact localization of cholesteatomas could be detected without fusion imaging using diffusion and T2-weighted magnetic resonance images in combination with computed tomography.

Methods

The study included patients with a diagnosis of cholesteatoma and had both temporal bone magnetic resonance imaging and computed tomography between 2017 and 2021. Presence of cholesteatomas was confirmed in non-echoplanar diffusion-weighted images prior to evaluation. Then, the localization of the lesion on computed tomography image was classified by detecting the equivalent of the lesion on T2-weighted thin-sliced images. All images were assessed by 2 independent radiologists. Kappa correlation coefficient was used to evaluate the interobserver agreement.

Results

Eighty-nine patients (49 female, 40 male) were included. The mean age was 39.07 (5-89). Interobserver agreement was moderate to almost perfect according to localization of cholesteatomas. The κ coefficient ranged between 0.48 and 0.83. The highest agreement was detected for the lesions located in the mesotympanium (κ=0.83), external acoustic canal, and mastoid bone. However, a moderate agreement was detected for lesions located in the medial epitympanic recess (κ=0.49).

Conclusion

Cholesteatomas which are placed in the mesotympanium, external auditory canal, and mastoid bone can be determined with high accuracy without fusion imaging.

Cite this article as

Perçem Orhan Söylemez U, Atalay B, Gündüz N, Gülçin Bozbeyoğlu S, Fırat A. Is fusion imaging mandatory to detect the localization of middle ear cholesteatomas? Inter-rater reliability in assessment. Current Research in MRI. 2022; 1(2): 42-46.