Developing an endscopic image based diagnosis system for the Otitis media disease

Abstract

Middle ear diseases diagnosis is highly dependent on the doctor’s experience thus some of the abnormalities can be easily missed even by experienced doctors. The process of reading medical images, detecting symptoms, and reaching a true conclusive diagnosis is a process that depends heavily on the doctor’s empirical knowledge, memory, intuition, and diligence. All of which are uncontrollable factors, which leaves the diagnostic process inconsistent, and prone to error for doctors who have been in the field for a while, and makes it extremely difficult for newly graduated doctors. This research aims to develop smart diagnosis system to support ENT (ear nose and throat) otolaryngologist that can identify and highlight lesions and regions of interests, and provide a second diagnosis opinion. For all patients fully Audiological and video otoscopic examination were done by the ENT otologists.