Cone-beam computed tomography (CBCT) is a popular imaging modality in dentistry for diagnosing and planning treatment for a variety of oral diseases with the ability to produce detailed, three-dimensional images of the teeth, jawbones, and surrounding structures. CBCT imaging has emerged as an essential diagnostic tool in dentistry. CBCT imaging has seen significant improvements in terms of its diagnostic value, as well as its accuracy and efficiency, with the most recent development of artificial intelligence (AI) techniques. This paper reviews recent AI trends and practices in dental CBCT imaging. AI has been used for lesion detection, malocclusion classification, measurement of buccal bone thickness, and classification and segmentation of teeth, alveolar bones, mandibles, landmarks, contours, and pharyngeal airways using CBCT images. Mainly machine learning algorithms, deep learning algorithms, and super-resolution techniques are used for these tasks. This review focuses on the potential of AI techniques to transform CBCT imaging in dentistry, which would improve both diagnosis and treatment planning. Finally, we discuss the challenges and limitations of artificial intelligence in dentistry and CBCT imaging.

翻译：锥形束计算机断层扫描（CBCT）是一种广泛应用于牙科的影像学检查方法，能够生成牙齿、颌骨及周围结构的精细三维图像，用于多种口腔疾病的诊断和治疗规划。CBCT 成像已成为牙科领域不可或缺的诊断工具。随着人工智能（AI）技术的最新发展，CBCT 成像在诊断价值、精度和效率方面均取得了显著提升。本文综述了近年来 AI 在牙科 CBCT 成像中的趋势与实践。AI 已用于基于 CBCT 图像的病变检测、错颌畸形分类、颊侧骨厚度测量，以及牙齿、牙槽骨、下颌骨、标志点、轮廓和咽部气道的分类与分割。主要采用机器学习算法、深度学习算法和超分辨率技术完成这些任务。本综述聚焦于 AI 技术变革牙科 CBCT 成像的潜力，这将同时改进诊断与治疗规划。最后，我们探讨了人工智能在牙科和 CBCT 成像中面临的挑战与局限。