Identifikasi Jenis Karies Gigi Menggunakan Algoritma You Only Look Once Versi 7 Berdasarkan Kedalaman Lubang

Abstract

Teeth are one of the organs in the human body that has a hard structure. The function of the teeth itself is for the process of chewing food so that it can be easily digested by the body. The part of the tooth itself consists of several parts, namely, the crown (the visible top), the root (the part embedded in the gum), dentin (hard tissue inside the tooth), enamel (a hard layer that protects the crown), and pulp (the inner part containing nerves and blood vessels). Teeth in general have several disease problems such as one of them is a tooth hole or what is called dental caries. Based on the depth of the tooth hole, caries can be divided into three, namely, superficial caries (tooth crown), media caries (tooth dentin) and profunda caries (tooth pulp). In general, it takes a lot of time and energy to diagnose the type of caries affected by a tooth. Therefore, a system is needed to be able to diagnose the type of caries affected in the teeth more quickly and efficiently. In this study, the You Only Look Once version 7 method is used to detect caries affected teeth based on the depth of the hole. The dataset used in the study was 857 caries images consisting of 748 training data and 109 validation data. The testing data used is 150 caries images which are divided into each class. The use of the YOLOv7 method to detect the type of caries based on the depth of the hole succeeded in getting an accuracy value of 86%.