Computer Assisted Classification of Dental Radiography

Dental radiographs have been extensively used in diagnosing dental diseases to help dentist for identifying victims of disasters. This study describes a flowchart of an automatic classification method for digital dental radiography. Thirty subjects were collected dental radiographs for this experiment. For each subject, there are fourteen different positions to take the X-ray images around the patient's mouth. A computer program was developed to automatically classify the fourteen classes of dental images, Upper Center Incisor (UCI), Lower Center Incisor (LCI), Upper Right Canine (URC), Upper Left Canine (ULC), Lower Right Canine (LRC), Lower Left Canine (LLC), Upper Right Premolar (URP), Upper Left Premolar (ULP), Lower Right Premolar (LRP), Lower Left Premolar (LLP), Upper Right Molar (URM), Upper Left Molar (ULM), Lower Right Molar (LRM) and Lower Left Molar (LLM). The classification method for digital dental radiography included image processing, feature computing and rule decisions for classifying. The results were compared with those of an experienced dentist. There were 420 images used to test the performance of proposed method. The accuracy of the classification is 80.5% (338/420).