Creation of automatic tooth classification algorithm using artificial intelligence and its application to identification system

2019 Fiscal Year Final Research Report

• Project/Area Number: 17K17113
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Pathobiological dentistry/Dental radiology
• Research Institution: Asahi University
• Principal Investigator: Nishiyama Wataru 朝日大学, 歯学部, 助教 (80631613)
• Project Period (FY): 2017-04-01 – 2020-03-31
• Keywords: 人工知能 / 歯式自動分類 / 歯科X線撮影 / パノラマX線画像

Outline of Final Research Achievements

A large number of unidentified persons occurred in the Great East Japan Earthquake, and dental information played a major role in identifying individuals. The collation of dental information was automated, but it required a large number of personnel to collect dental information from unidentified bodies. We have developed an algorithm that uses artificial intelligence "Deep Learning" to automatically extract dental information from a panoramic radiography, which is a general image in the dental field. Three algorithms have been developed: (1) detection of a single tooth, (2) automatic tooth type classification, and (3) tooth condition classification. (1) Sensitivity 96.4% for detection of single tooth, (2) Accuracy 93.2% for automatic tooth type classification, (3) Sensitivity 98.0% for tooth condition classification, indicating the possibility of automatic extraction of dental information.

Free Research Field

歯科放射線学

Academic Significance and Societal Importance of the Research Achievements

パノラマから歯科情報を自動抽出が可能となれば緊急時に必要となる医療資源の削減につながり、照合を早期に行なうことにより経時的な遺体の劣化による個人特定のための情報消失も防ぐことができる。また医療施設のパノラマX線撮影装置に自動抽出システムを採用することにより平時の恒常的な歯科情報蓄積が可能となり、身元不明遺体の生前情報の採取が容易になる。