Private practice, Saint Quentin, France.
Private practice, Como, Italy.
Eur J Orthod. 2023 Nov 30;45(6):818-831. doi: 10.1093/ejo/cjad060.
Cone-beam computed tomography (CBCT) has several applications in various fields of dental medicine such as diagnosis and treatment planning. When compared to computed tomography (CT), CBCT's radiation exposure dose is decreased by 3%-20%. However, CBCT produces more scattered signals and may present poorer image quality when compared to medical CT.
To review the findings regarding the accuracy of multi-detector computed tomography (MDCT) and CBCT and to compare the different software programs that segment the upper airway.
Three databases (PubMed, Medline, and Web of Science) were searched for articles and a manual search was performed.
The inclusion criteria were defined following the PICO framework: P-any patient with a CBCT or CT; I-dimensional evaluation of the upper airway using MDCT or CBCT; C-phantoms; O-the primary outcome was MDCT and CBCT accuracy, the secondary outcome was the evaluation and comparison of software programs used to segment the upper airway.
Articles that met eligibility criteria were assessed using the Critical Appraisal Skills Program Checklist.
Among the 16 eligible studies, 6 articles referred to the accuracy of MDCTs or CBCTs and 10 to the accuracy of the software. Most articles were qualified as high quality.
MDCT and CBCT scans' accuracy in upper airway dimensional measurements depends on machine brand, parameters, and segmentation technique. Regarding the segmentation technique, 12 programs were studied. Most either underestimated or overestimated upper airway measurements. In particular, OnDemand3D and INVIVO showed poor accuracy. On the contrary, Invesalius, and MIMICS were accurate in assessing nasal cavities when using an interactive threshold. However, results varied due to methodological differences among the studies. Finally, fully automatic segmentation based on artificial intelligence may represent the future of airway segmentation because it is faster and seems to be accurate. However, further studies are necessary.
This study was registered in Prospero (International Prospective Register of Systematic Reviews) with the ID number CRD42022373998.
锥形束计算机断层扫描（CBCT）在口腔医学的诊断和治疗规划等多个领域有多种应用。与计算机断层扫描（CT）相比，CBCT 的辐射剂量降低了 3%-20%。然而，与医学 CT 相比，CBCT 产生的散射信号更多，图像质量可能较差。
综述多探测器 CT（MDCT）和 CBCT 的准确性，并比较用于分割上气道的不同软件程序。
在 PubMed、Medline 和 Web of Science 三个数据库中检索文章，并进行手动搜索。
纳入标准遵循 PICO 框架：P-任何接受 CBCT 或 CT 的患者；I-使用 MDCT 或 CBCT 对气道进行一维评估；C-体模；O-主要结局为 MDCT 和 CBCT 的准确性，次要结局为用于分割上气道的软件程序的评估和比较。
符合纳入标准的文章使用批判性评估技能计划清单进行评估。
在 16 篇符合条件的研究中，有 6 篇文章涉及 MDCT 或 CBCT 的准确性，10 篇文章涉及软件的准确性。大多数文章质量较高。
MDCT 和 CBCT 对上气道尺寸测量的准确性取决于机器品牌、参数和分割技术。关于分割技术，研究了 12 个程序。大多数程序要么低估，要么高估上气道的测量值。特别是，OnDemand3D 和 INVIVO 显示出较差的准确性。相反，使用交互式阈值时，Invesalius 和 MIMICS 对上鼻腔的评估较为准确。然而，由于研究之间的方法学差异，结果有所不同。最后，基于人工智能的全自动分割可能代表气道分割的未来，因为它更快，而且似乎更准确。然而，还需要进一步的研究。
本研究在 Prospero（国际前瞻性系统评价注册库）中以注册号 CRD42022373998 进行了注册。
