Early diagnosis for the onset of peri-implantitis based on artificial neural network

Wanting Fan; Jianming Tang; Huixia Xu; Xilin Huang; Donglei Wu; Zheng Zhang

doi:10.1515/biol-2022-0691

Early diagnosis for the onset of peri-implantitis based on artificial neural network

Open Life Sci. 2023 Aug 31;18(1):20220691. doi: 10.1515/biol-2022-0691. eCollection 2023.

Authors

Wanting Fan¹, Jianming Tang¹, Huixia Xu¹, Xilin Huang², Donglei Wu¹, Zheng Zhang¹

Affiliations

¹ Department of Stomatology, Shenzhen People's Hospital, Shenzhen, Guangdong, China.
² Department of Obstetrics, Shenzhen People's Hospital, Shenzhen, Guangdong, China.

Abstract

The aim of this study is to construct an artificial neural network (ANN) based on bioinformatic analysis to enable early diagnosis of peri-implantitis (PI). PI-related datasets were retrieved from the Gene Expression Omnibus database. Differentially expressed genes (DEGs) and functional enrichment analyses were performed between PI and the control group. Furthermore, the infiltration of 22 immune cells in PI was analyzed using CIBERSORT. Hub genes were identified with random forest (RF) classification. The ANN model was then constructed for early diagnosis of PI. A total of 1,380 DEGs were identified. Enrichment analysis revealed the involvement of neutrophil-mediated immunity and the NF-kappa B signaling pathway in PI. Additionally, higher proportion of naive B cells, activated memory CD4 T cells, activated NK cells, M0 macrophages, M1 macrophages, and neutrophils were observed in the soft tissues surrounding PI. From the RF analysis, 13 hub genes (ST6GALNAC4, MTMR11, SKAP2, AKR1B1, PTGS2, CHP2, CPEB2, SYT17, GRIP1, IL10, RAB8B, ABHD5, and IGSF6) were selected. Subsequently, the ANN model for early diagnosis of PI was constructed with high performance. We identified 13 hub genes and developed an ANN model that accurately enables early diagnosis of PI.

Keywords: artificial neural network; early diagnosis; machine learning; peri-implantitis.