ABSTRACT
Background/purpose: Proper implant-ridge classification is crucial for developing a dental implant treatment plan. This study aimed to verify the ability of MobileNet, an advanced deep learning model characterized by a lightweight architecture that allows for efficient model deployment on resource-constrained devices, to identify the implant-ridge relationship. Materials and methods: A total of 630 cone-beam computerized tomography (CBCT) slices from 412 patients were collected and manually classified according to Terheyden's definition, preprocessed, and fed to MobileNet for training under the conditions of limited datasets (219 slices, condition A) and full datasets (630 cases) without and with automatic gap filling (conditions B and C). Results: The overall model accuracy was 84.00% in condition A and 95.28% in conditions B and C. In condition C, the accuracy rates ranged from 94.00 to 99.21%, with F1 scores of 89.36-100.00%, and errors due to unidentifiable bone-implant contact and miscellaneous reasons were eliminated. Conclusion: The MobileNet architecture was able to identify the implant-ridge classification on CBCT slices and can assist clinicians in establishing a reliable preoperative diagnosis and treatment plan for dental implants. These results also suggest that artificial intelligence-assisted implant-ridge classification can be performed in the setting of general dental practice.
ABSTRACT
OBJECTIVES: This study aimed to use a deep learning (DL) approach for the automatic identification of the ridge deficiency around dental implants based on an image slice from cone-beam computerized tomography (CBCT). MATERIALS AND METHODS: Single slices crossing the central long-axis of 630 mandibular and 845 maxillary virtually placed implants (4-5 mm diameter, 10 mm length) in 412 patients were used. The ridges were classified based on the intraoral bone-implant support and sinus floor location. The slices were either preprocessed by alveolar ridge homogenizing prior to DL (preprocessed) or left unpreprocessed. A convolutional neural network with ResNet-50 architecture was employed for DL. RESULTS: The model achieved an accuracy of >98.5% on the unpreprocessed image slices and was found to be superior to the accuracy observed on the preprocessed slices. On the mandible, model accuracy was 98.91 ± 1.45%, and F1 score, a measure of a model's accuracy in binary classification tasks, was lowest (97.30%) on the ridge with a combined horizontal-vertical defect. On the maxilla, model accuracy was 98.82 ± 1.11%, and the ridge presenting an implant collar-sinus floor distance of 5-10 mm with a dehiscence defect had the lowest F1 score (95.86%). To achieve >90% model accuracy, ≥441 mandibular slices or ≥592 maxillary slices were required. CONCLUSIONS: The ridge deficiency around dental implants can be identified using DL from CBCT image slices without the need for preprocessed homogenization. The model will be further strengthened by implementing more clinical expertise in dental implant treatment planning and incorporating multiple slices to classify 3-dimensional implant-ridge relationships.
