Establishment of a system for predicting the development of peri-implantitis using artificial intelligence.

2023 Fiscal Year Final Research Report

• Project/Area Number: 21K17040
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 57050:Prosthodontics-related
• Research Institution: Osaka University
• Principal Investigator: Nishimura Yuichi 大阪大学, 大学院歯学研究科, 招へい教員 (70883263)
• Project Period (FY): 2021-04-01 – 2024-03-31
• Keywords: インプラント周囲炎 / 予測

Outline of Final Research Achievements

The purpose of this study was to establish a system to predict the onset of peri-implantitis with high accuracy. A support vector machine (SVM) was used as a machine learning model, and variables used in the analysis were factors related to the development of peri-implantitis selected from previous reports. After randomly adjusting the number of samples in the normal and periarthritis groups to be equal, we constructed a training model using 70% of the adjusted samples as training data and predicted the onset of periarthritis using the remaining 30% as validation data, and compared the predicted values obtained with the actual values. The accuracy of prediction by SVM was 0.79, the goodness of fit was 0.73, the reproducibility was 0.88, the F value was 0.80, and the AUC was 0.81.

Free Research Field

歯科補綴学

Academic Significance and Societal Importance of the Research Achievements

本研究においてインプラント体の周囲骨吸収は，過去に関連が報告されている糖尿病，口腔清掃状態や角化粘膜幅などの炎症誘発性因子に加え，人工知能を用いた分析を行うことで固定様式，上顎骨への埋入といった構造や解剖学的因子，さらには補綴学的因子との関連が示された．長期経過したインプラント体を対象とした周囲骨吸収のリスク因子を検討した研究は少なく，本研究から得られた結果は、インプラント周囲炎の発症予測システムの確立基盤となり得ることから臨床的意義は大きいと考えられる．