The Design of Convolutional Neural Networks Model for Classification of Ear Diseases on Android Mobile Devices

An otorhinolaryngologist (ORL) or general practitioner diagnoses ear disease based on image information. However, practitioners refer patients to ORL for chronic because the of has high complexity, variety, and little difference between diseases. artificial intelligence-based approach is needed make it easier doctors diagnose diseases information, such as Convolutional Neural Network (CNN). This paper describes how CNN was designed generate models used classify The model developed using an dataset from practice at University Mataram Teaching Hospital. work aims find best classifying applicable android mobile devices. Furthermore, deployed Android-based application integrated with Endoscope Ear Cleaning Tool Kit registering patient images. experimental results show 83% accuracy, 86% precision, recall, 4ms inference time. produces a System Usability Scale 76.88% testing, which shows easy use. achievement that can be into ENT expert system. In future, system operated by workers in community health centres/clinics assist leading them diagnosing early.