Progress in deep learning-based dental and maxillofacial image analysis: A systematic review

Background: With the advent of deep learning in modern computing there has been unprecedented progress in image processing and segmentation. Deep learning-based image pattern recognition achieved a significant place in interpreting dental radiographs towards automatic diagnosis and treatment. In context with dental imaging, deep learning-based image analysis has been able to perform dental structure segmentation, classification, and identification of several common dental diseases with significant 90% accuracy. These results open a window of hope for better diagnosis and treatment planning in dental medicine. This review systematically presents recent advances in deep learning-based dental and maxillofacial image analysis. Materials and methods: We performed an extensive literature survey using the PubMed literature repository for identifying suitable articles. We shortlisted more than 75 articles that use deep learning for dental image seg-mentation, object detection, classification, and other image processing-related tasks. This study includes vari-ables such as the size of the dataset, dental imaging modality, deep learning architecture, and performance evaluation measures. Results: We have summarized recent developments and a concise overview of studies on various applications of dental and maxillofacial image analysis. We primarily discussed how deep learning techniques have been exploited in areas such as tooth detection and labeling, dental caries, plaque, periodontal condition, osteoporosis, oral lesion, anatomical landmarking, age, and gender estimation. The challenges and future research directions in the area have been extensively discussed. Conclusion: Undoubtedly remarkable progress is witnessed in dental image analysis in recent years. However, many crucial aspects still need to be addressed including standardization of data and generalization in AI-based solutions towards dental and maxillofacial image analysis for the diagnosis and better treatment aid in the field of dentistry which will open a new avenue in dental clinical practices.