华西口腔医学杂志

• 临床研究 • 上一篇    下一篇

计算机辅助的牙槽骨密度定量测量系统准确性和灵敏度的研究

郑旭1 林久祥2   

  1. 1.北京大学第三医院口腔科, 北京100191; 2.北京大学口腔医院正畸科, 北京100081
  • 收稿日期:2009-06-25 修回日期:2009-06-25 出版日期:2009-06-20 发布日期:2009-06-20
  • 通讯作者: 郑旭,Tel:010-82266338
  • 作者简介:郑旭(1973-),女,河北人,副主任医师,博士

Accuracy and sensibility of computer -aided alveolar bone density quantitative measuring system

ZHENG Xu1, LIN Jiu-xiang2   

  1. 1. Dept. of Stomatology, The Third Hospital of Peking University, Beijing 100191, China; 2. Dept. of Orthodontics, School of Stomatology, Peking University, Beijing 100081, China
  • Received:2009-06-25 Revised:2009-06-25 Online:2009-06-20 Published:2009-06-20
  • Contact: ZHENG Xu,Tel:010-82266338

摘要:

目的建立计算机辅助的牙槽骨密度定量测量系统,并通过样本测试检验系统的测量准确性及灵敏度,实现对牙槽骨中羟磷灰石(HP)含量的估测。方法在X线骨密度测量法基础上利用定位投照的直接数字化牙片,以铝梯为参照校正投照条件差异导致的影像灰度误差,辅以计算机技术建立牙槽骨密度定量测量系统。用该系统测量HP含量不同的19个样本的平均灰度值,根据已知的铝梯灰度和厚度推算这些灰度值对应的等效铝厚度和校正灰度。建立等效铝厚度、校正灰度与HP含量之间的一元线性回归方程,将等效铝厚度、校正灰度回代入方程计算出估测的HP含量;与HP的实际含量相比,计算估测的偏误率。随机选择2个样本重复测量10次,计算牙槽骨密度定量测量系统的测量误差范围,结合回归方程,计算该系统检测HP含量变化的灵敏度。结果等效铝厚度、校正灰度与HP含量之间有明显相关性;当测得的等效铝厚度大于0.67 mm或校正灰度大于41个灰阶时,对HP含量估测的偏误率在5%以下。用等效铝厚度进行估测时,HP含量的变化大于0.17 mg/mm2即可被有效检出;用校正灰度进行估测时,HP含量的变化大于0.18 mg/mm2即可被有效检出。结论本研究牙槽骨密度定量测量系统的准确性和灵敏度能够满足临床和科研的需要,可用于对牙槽骨密度的横向比较和纵向观察,有较广阔的应用空间。

关键词: 牙槽骨, 骨密度, 数字化牙片, 准确性, 灵敏度

Abstract:

Objective The alveolar bone density quantitative measuring system was established on the basis of grey level measuring on standardized-exposed direct digital periapical radiograph by referring to an aluminum step wedge with the aid of computer technique. We tested it′s accuracy and sensibility by measuring a series of specimens and made it possible to estimate the area density of hydroxyl-phosphorite(HP) in the alveolar bone with the system. Methods The average grey level intensity of 19 specimens containing different amount of HP within certain area was measured with the system. The equivalent aluminum thickness(EAT) and corrected grey level intensity(CGL) of each specimen were calculated according to known grey level intensity of the aluminum step wedge and it′ s corresponding thickness. The liner regression equations between EAT, CGL and HP density(mg/mm2) were established. We put the values of EAT and CGL into the equations and calculated the corresponding values of estimated HP density. The bias ratios of estimation were then calculated. We measured two randomly chosen specimens 10 times respectively to calculate the system′s measurement error range, then the sensibility of the system for measuring HP density was calculated according to the regression equations. Results EAT and CGL were significantly correlated with HP density. When EAT was larger than 0.67 mm or CGL larger than 41, the system′s bias ratio of estimated HP density was lower than 5%. When we estimated the HP density with the use of EAT, the system could effectively check out a HP density change larger than 0.17 mg/mm2, with the use of CGL, a HP density change larger than 0.18 mg/mm2 could be effectively checked out. Conclusion The accuracy and sensibility of the quantitative alveolar bone density measuring system are acceptable, it can be widely used to compare alveolar bone density longitudinally and cross-sectionally.

Key words: alveolar bone, bone density, digital periapical radiograph, accuracy, sensibility