Comparison of Different Slice Thicknesses in Computed Tomography for Labyrinthine Fistula Evaluation.

ABSTRACT

OBJECTIVES: The aim of the present study is to assess the impact of different slice thicknesses in computed tomography for labyrinthine fistula evaluation and to determine the appropriate slice thickness. METHODS: A total of 258 patients who underwent mastoidectomy and tympanoplasty for chronic otitis media with cholesteatoma from 2010 to 2014 were reviewed. The radiological diagnoses were compared with intraoperative findings. Sensitivity and specificity of 2.0-, 1.5-, 1.0-, and 0.75-mm-thick computed tomographic (CT) images for the evaluation of labyrinthine fistulae were calculated. Cohen's κ coefficient was also calculated. RESULTS: The sensitivities of 2.0-, 1.5-, 1.0-, and 0.75-mm-thick CT images for the evaluation of labyrinthine fistulae were 76.9, 86.5, 90.4, and 93.3% (observer 1) and 67.3, 76.0, 79.8, and 87.5% (observer 2), respectively. The specificities of 2.0-, 1.5-, 1.0-, and 0.75-mm-thick CT images for labyrinthine fistula evaluation were 96.1, 94.8, 95.5, and 95.5% (observer 1) and 99.4, 97.4, 95.5, and 94.8% (observer 2), respectively. Cohen's κ coefficients were 0.790, 0.788, 0.876, and 0.911 in 2.0-, 1.5-, 1.0-, and 0.75-mm-thick CT images, respectively. CONCLUSIONS: The sensitivity of CT for labyrinthine fistula evaluation increases with decreasing slice thickness, while the specificity does not improve.