RESUMO
OBJECTIVES: The presence of insufficient peri-implant supracrestal tissue height (STH) may increase marginal bone resorption. This study aims to evaluate the effect of STH on marginal bone level changes (ΔMBC) in platform-switching posterior implants placed crestally and subcrestally. METHODS: A total of 80 implants were included in this study. There were two main groups in the study; STH≤2 mm (A) and STH> 2 mm (B) and four subgroups according to the implant placement level, crestally (I) and subcrestally (II): A-I, A-II, B-I, and B-II. Intraoperatively, STH and placement depths of implants were measured from mesial and distal aspects. The mesial and distal peri-implant marginal bone levels were measured on periapical radiographs at immediately (T0), 6 months (T1), 9 months (T2), and 12 months (T3) after functional loading, and the difference between the marginal bone levels was calculated as the ΔMBC. RESULTS: Statistically significantly more mesial ΔMBC was detected in the A-I than in the B-I at the time of T0-T1. In the group with STH greater than 2 mm, the difference in ΔMBC between the crestally and subcrestally placements was not statistically significant. CONCLUSIONS: This study was found that STH is effective at protecting the marginal bone in the early period, and in cases where the STH is insufficient, subcrestally placement may increase long-term implant success by preventing marginal bone loss from occurring beyond the implant shoulder. The clinical trial number is NCT05595746. CLINICAL SIGNIFICANCE: In this study, it was demonstrated that an STH greater than 2 mm is important for marginal bone stabilization, regardless of crestal and subcrestal levels, and that subcrestally placement is beneficial in cases of insufficient STH.
RESUMO
OBJECTIVES: The objective of this study is to assess the accuracy of computer-assisted periodontal classification bone loss staging using deep learning (DL) methods on panoramic radiographs and to compare the performance of various models and layers. METHODS: Panoramic radiographs were diagnosed and classified into 3 groups, namely "healthy," "Stage1/2," and "Stage3/4," and stored in separate folders. The feature extraction stage involved transferring and retraining the feature extraction layers and weights from 3 models, namely ResNet50, DenseNet121, and InceptionV3, which were proposed for classifying the ImageNet dataset, to 3 DL models designed for classifying periodontal bone loss. The features obtained from global average pooling (GAP), global max pooling (GMP), or flatten layers (FL) of convolutional neural network (CNN) models were used as input to the 8 different machine learning (ML) models. In addition, the features obtained from the GAP, GMP, or FL of the DL models were reduced using the minimum redundancy maximum relevance (mRMR) method and then classified again with 8 ML models. RESULTS: A total of 2533 panoramic radiographs, including 721 in the healthy group, 842 in the Stage1/2 group, and 970 in the Stage3/4 group, were included in the dataset. The average performance values of DenseNet121 + GAP-based and DenseNet121 + GAP + mRMR-based ML techniques on 10 subdatasets and ML models developed using 2 feature selection techniques outperformed CNN models. CONCLUSIONS: The new DenseNet121 + GAP + mRMR-based support vector machine model developed in this study achieved higher performance in periodontal bone loss classification compared to other models in the literature by detecting effective features from raw images without the need for manual selection.
