Transfer Learning via Deep Neural Networks for Implant Fixture System Classification Using Periapical Radiographs

Abstract

In the absence of accurate medical records, it is critical to correctly classify implant fixture systems using periapical radiographs to provide accurate diagnoses and treatments to patients or to respond to complications. The purpose of this study was to evaluate whether deep neural networks can identify four different types of implants on intraoral radiographs. In this study, images of 801 patients who underwent periapical radiographs between 2005 and 2019 at Yonsei University Dental Hospital were used. Images containing the following four types of implants were selected: Brånemark Mk TiUnite, Dentium Implantium, Straumann Bone Level, and Straumann Tissue Level. SqueezeNet, GoogLeNet, ResNet-18, MobileNet-v2, and ResNet-50 were tested to determine the optimal pre-trained network architecture. The accuracy, precision, recall, and F1 score were calculated for each network using a confusion matrix. All five models showed a test accuracy exceeding 90%. SqueezeNet and MobileNet-v2, which are small networks with less than four million parameters, showed an accuracy of approximately 96% and 97%, respectively. The results of this study confirmed that convolutional neural networks can classify the four implant fixtures with high accuracy even with a relatively small network and a small number of images. This may solve the inconveniences associated with unnecessary treatments and medical expenses caused by lack of knowledge about the exact type of implant.

Deep Neural Networks for Dental Implant System Classification

“…The same was found to be true for the classification of dental implants using CNNs [13]. From our study, we found that CNNs using panoramas lead to results comparable to the diagnostic accuracy of dental implant classification by CNNs using periapical radiographic images [13]. These results will contribute to the accuracy of CNN classification diagnoses by increasing the number of images used via preprocessing.…”

“…Because panoramic radiographs have different distortions depending on the region to be photographed, periapical radiographic images have generally been used for diagnosis, whereas CNNs have been used for tooth-related classifications and diagnoses [32,33]. The same was found to be true for the classification of dental implants using CNNs [13]. From our study, we found that CNNs using panoramas lead to results comparable to the diagnostic accuracy of dental implant classification by CNNs using periapical radiographic images [13].…”

“…However, doing so requires a significant amount of human effort and experience. No automated method has yet been proposed for identifying implant brands from panoramic radiograph images [13]. Clinically, this capability would be useful.…”

Artificial Intelligence in Fractured Dental Implant Detection and Classification: Evaluation Using Dataset from Two Dental Hospitals

“…In particular, as newly developed DCNN models and algorithms are continuously adopted and coupled with the area of implant dentistry, it may be an important adjunct for diagnosis, treatment, and prognosis assessments [ 26 ]. Recent DCNN-related studies confirmed that various types of DIs with different shapes, lengths, or dimensions can be effectively detected and classified using panoramic and periapical images [ 27 , 28 , 29 ].…”

A Model Transfer Learning Framework With Back-Propagation Neural Network for Wine and Chinese Liquor Detection by Electronic Nose

“…With the development of machine learning, model-based transfer learning has become a popular method in the field of computer vision because it allows for accurate modeling and takes less time. It is also effective in applying learned features from large datasets to small datasets to raise their accuracy and performance [26]. With transfer learning, the model to be trained does not start from scratch, but from the patterns that have been learned when solving various problems.…”