Classification of Tympanic Membrane Images based on VGG16 Model

Otitis Media (OM) is a type of infectious disease caused by viruses and/or bacteria in the middle ear cavity. In current study, it aimed to detect eardrum region images for diagnosing OM using artificial intelligence methods. The Convolution Neural Networks (CNN) model and deep features this obtained with otoscope device were used. order separate these as Normal Abnormal, end-to-end VGG16 was directly used first stage experimental work. second activation maps fc6 fc7 layers consisting 4096 fc8 layer 1000 CNN obtained. Then, given input Support Vector Machines (SVM). from all combined new feature set last stage, an SVM. Thus, effect on success distinguishing investigated. Experimental studies show that, best performance results accuracy rate 82.17%. addition, 71.43%, 90.62% 77.92% criteria sensitivity, specificity f-score values, respectively. Consequently, has been shown that could be accurately detected architecture. proposed learning-based classification system promises highly accurate detection.