TVGG Dental Implant Identification System

Jianbin Guo; Pei-Wei Tsai; Xingsi Xue; Dong Wu; Qui Tran Van; Chanaka Nimantha Kaluarachchi; Hong Thi Dang; Nikhitha Chintha

doi:10.3389/fphar.2022.948283

TVGG Dental Implant Identification System

Front Pharmacol. 2022 Aug 8:13:948283. doi: 10.3389/fphar.2022.948283. eCollection 2022.

Authors

Jianbin Guo¹, Pei-Wei Tsai², Xingsi Xue³, Dong Wu⁴, Qui Tran Van², Chanaka Nimantha Kaluarachchi², Hong Thi Dang², Nikhitha Chintha²

Affiliations

¹ Fujian Key Laboratory of Oral Diseases and Fujian Provincial Engineering Research Center of Oral Biomaterial and Stomatological Key Lab of Fujian College and University, School and Hospital of Stomatology, Fujian Medical University, Fuzhou, China.
² Department of Computing Technologies, Swinburne University of Technology, Hawthorn, VIC, Australia.
³ Fujian Provincial Key Laboratory of Big Data Mining and Applications, Fujian University of Technology, Fuzhou, China.
⁴ Institute of Stomatology and Research Center of Dental and Craniofacial Implants, School and Hospital of Stomatology, Fujian Medical University, Fuzhou, China.

Abstract

Identifying the right accessories for installing the dental implant is a vital element that impacts the sustainability and the reliability of the dental prosthesis when the medical case of a patient is not comprehensive. Dentists need to identify the implant manufacturer from the x-ray image to determine further treatment procedures. Identifying the manufacturer is a high-pressure task under the scaling volume of patients pending in the queue for treatment. To reduce the burden on the doctors, a dental implant identification system is built based on a new proposed thinner VGG model with an on-demand client-server structure. We propose a thinner version of VGG16 called TVGG by reducing the number of neurons in the dense layers to improve the system's performance and gain advantages from the limited texture and patterns in the dental radiography images. The outcome of the proposed system is compared with the original pre-trained VGG16 to verify the usability of the proposed system.

Keywords: CNN; VGG; dental implant; implant identification; radiography image.