目的：探讨通过锥形束CT（cone-beam computed tomography, CBCT）评判上颌前牙骨开裂和骨开窗的真实性和可靠性。方法：纳入18例安氏Ⅲ类错牙合行骨皮质切开术的患者，平均年龄23.6岁（18~30岁）， 其中男3例，女15例，观察108颗上颌前牙。以翻瓣术中所见的骨开裂和骨开窗作为判定金标准，评价CBCT判断骨开裂和骨开窗的灵敏度（sensitivity）、特异度（specificity）、约登指数（Youden index）、阳性似然比（positive likelihood ratio）、阴性似然比（negative likelihood ratio）、阳性预测值（positive predictive value）和阴性预测值（negative predictive value），并将CBCT判定与翻瓣直视判定相比较。结果：上颌前牙骨开裂和骨开窗的发生率分别为10.19%和13.89%，主要见于侧切牙和尖牙。骨开裂长度中位数为5 mm，宽度中位数为4 mm；骨开窗常位于根中部至根尖部，长度中位数为3 mm，宽度中位数为2 mm。骨开裂CBCT判定与临床判定之间有较好的一致性（P<0.05）， CBCT判定的灵敏度和特异度均>0.7。骨开窗CBCT判定与临床判定之间一致性一般（P<0.05）， 灵敏度为0.93，特异度为0.52。结论：CBCT评判上前牙骨开裂与临床情况的一致性较好，判定骨开窗与临床情况的一致性一般。CBCT判定骨开裂与骨开窗在临床中有一定应用价值，但存在局限性。

Objective：To evaluate the accuracy and reliability of detecting alveolar bone dehiscence and fenestration of maxillary anterior teeth of Angle class Ⅲ by cone-beam computed tomography (CBCT). Methods: Eighteen Angle class Ⅲ patients with 108 maxillary anterior teeth were included (3 males and 15 females) who accepted modified corticotomy in orthodontic therapy. The mean age was 23.6 years (18-30 years). The clinical detection of dehiscence and fenestration was done when modified corticotomy was performed by the same periodontist. The CBCT examination was conducted pre-ope-ration and the detection of dehiscence and fenestration by CBCT was done by two periodontists. The data in modified corticotomy were used as the golden standard to calculate the parameters, such as sensitivity, specificity, positive and negative predictive values, Youden index (YI), positive and negative likelihood ratio. Kappa statistic was used to analyze the agreement between the clinical detection and the CBCT detection. Results: The incidence of dehiscence and fenestration was about 10.19% and 13.89% respectively, which mainly occurred on lateral incisors and canines. The median values of length and width of dehiscence were about 5 mm and 4 mm, and the median values of length and width of fenestration were 3 mm and 2 mm, respectively. Most fenestrations were detected on the middle third to the apical third of the root. For dehiscence, the agreement between clinical detection and CBCT detection was statistically significant (P<0.05). For fenestration, the agreement between clinical detection and CBCT detection was statistically significant (P<0.05). The values of sensitivity and specificity for detecting dehiscence were more than 0.7. The values of positive and negative predictive values for detecting dehiscence were 0.44 and 0.97. The values of sensitivity and specificity for detecting fenestration were 0.93 and 0.52. The values of positive and negative predictive values for detecting fenestration were 0.24 and 0.98. Conclusion: For dehiscence, the agreement between clinical detection and CBCT detection was good. For fenestration, the agreement between clinical detection and CBCT detection was general. Detection of dehiscence and fenestration of maxillary anterior teeth of Angle class Ⅲ by CBCT had limited diagnostic value in clinical practice with overestimation of dehiscence and fenestration incidence.