A Performance Comparison between Automated Deep Learning and Dental Professionals in Classification of Dental Implant Systems from Dental Imaging: A Multi-Center Study.

Jae-Hong Lee,Young-Taek Kim,Jong-Bin Lee,Seong-Nyum Jeong

doi:10.3390/diagnostics10110910

Jae-Hong Lee, Young-Taek Kim + Show 2 more

Open Access

https://doi.org/10.3390/diagnostics10110910

Copy DOI

Abstract

In this study, the efficacy of the automated deep convolutional neural network (DCNN) was evaluated for the classification of dental implant systems (DISs) and the accuracy of the performance was compared against that of dental professionals using dental radiographic images collected from three dental hospitals. A total of 11,980 panoramic and periapical radiographic images with six different types of DISs were divided into training (n = 9584) and testing (n = 2396) datasets. To compare the accuracy of the trained automated DCNN with dental professionals (including six board-certified periodontists, eight periodontology residents, and 11 residents not specialized in periodontology), 180 images were randomly selected from the test dataset. The accuracy of the automated DCNN based on the AUC, Youden index, sensitivity, and specificity, were 0.954, 0.808, 0.955, and 0.853, respectively. The automated DCNN outperformed most of the participating dental professionals, including board-certified periodontists, periodontal residents, and residents not specialized in periodontology. The automated DCNN was highly effective in classifying similar shapes of different types of DISs based on dental radiographic images. Further studies are necessary to determine the efficacy and feasibility of applying an automated DCNN in clinical practice.

Highlights

Dental implants have become a predictable treatment alternative for patients with partial or complete edentulous conditions [1]
The accuracy of the automated deep convolutional neural network (DCNN) abased on the area under the ROC curve (AUC), Youden index, sensitivity, and specificity for the 2,396 panoramic and periapical radiographic images were 0.954, 0.808, 0.955, and 0.853, respectively
Using only panoramic radiographic images (n = 1429), the automated DCNN achieved an AUC of 0.929, while the corresponding value using only periapical radiographic images (n = 967) achieved an AUC of 0.961

Summary

Introduction

Dental implants have become a predictable treatment alternative for patients with partial or complete edentulous conditions [1]. Over the years, this treatment modality has evolved as a standard treatment protocol for replacing missing teeth. Hundreds of manufacturers worldwide are producing and distributing over 2000 different types of dental implant systems (DISs) that differ in diameter, length, shape, coating, and surface material and properties [2,3]. DISs have shown a success rate of more than 90% and long-term survival rate of more than 10 years in systematic and meta-analytic review studies, which inevitably increases with the occurrence of mechanical and biological complications, such as fixtures or screw fractures, screw loosening, veneer chipping or fractures, low implant stability, peri-implant mucositis, and peri-implantitis [4,5,6,7].

Objectives

Methods

Results

Discussion

Conclusion