Asymmetry Computing for Cholesteatoma Detection Based on 3-D CT Images

摘要

Cholesteatoma is a destructive and expanding sac in the middle ear and/or mastoid process. If untreated, cholesteatoma can result in nerve deterioration, deafness, imbalance and vertigo. Traditional diagnose methods rely on doctors’ experience and often misdiagnosed. In this paper, we propose a novel asymmetry-computing algorithm for cholesteatoma detection based on 3-D CT images. By applying this algorithm, we provide a complete numerical calculation framework and its simulation. The proposed algorithm is tested on real 3-D eardrum CT images. The diagnose accuracy rate of cholesteatoma is 72.73%, and the misdiagnose rate is only 27.27%. The result demonstrates the applicability of the proposed asymmetry-computing algorithm for cholesteatoma detection. Therefore the proposed algorithm is beneficial for clinic diagnose.