Suppr超能文献

开发并验证一种实用的 EOS X 射线系统运动伪影检测解决方案。

Development and validation of a practical solution for detecting motion artefacts in the EOS X-ray system.

机构信息

Department of Radiology, Swiss Paraplegic Centre, Nottwil, CH, Switzerland.

Swiss Paraplegic Research, Nottwil, CH, Switzerland.

出版信息

Sci Rep. 2024 Feb 28;14(1):4837. doi: 10.1038/s41598-024-55373-2.

Abstract

The EOS™2D/3D system is a low-dose, 3D imaging system that utilizes two perpendicular X-ray beams to create simultaneous frontal and lateral images of the body. This is a useful modality to assess spinal pathologies. However, due to the slow imaging acquisition time up to 25 s, motion artifacts (MA) frequently occur. These artifacts may not be distinguishable from pathological findings, such as scoliosis, and may impair the diagnostic process. The aim of this study was to design a method to detect MA in EOS X-ray. We retrospectively analyzed EOS imaging from 40 patients wearing a radiopaque reference device during imaging. We drew a straight vertical line along the reference device. We measured deviations from it to quantify MA, presenting these findings through descriptive statistics. For a subset of patients with high MA, acquisitions were repeated after giving specific instructions to stand still. For these patients, we compared MA between the two acquisitions. In our study, a substantial proportion of patients exhibited MA ≥ 1 mm, with 80% in frontal projections and 87.9% in lateral projections. In the subjects who received a second acquisition, MA was significantly lower in the second images. Our method allows for a precise detection of MA on EOS images through a simple, yet reliable solution. Our method may improve the reliability of spine measurements, and reduce the risk of wrong diagnosis due to low imaging quality.

摘要

EOS™2D/3D 系统是一种低剂量的 3D 成像系统,它利用两个垂直的 X 射线束同时创建身体的正面和侧面图像。这是评估脊柱病变的一种有用方法。然而,由于成像采集时间长达 25 秒,运动伪影(MA)经常发生。这些伪影可能与脊柱侧凸等病理发现难以区分,并可能影响诊断过程。本研究旨在设计一种检测 EOS X 射线中 MA 的方法。我们回顾性分析了 40 名患者在成像时佩戴不透射线参考装置的 EOS 成像。我们沿着参考装置画了一条垂直的直线。我们测量了与它的偏差,以量化 MA,并通过描述性统计来呈现这些发现。对于 MA 较高的患者亚组,在给予站立不动的具体说明后重复采集。对于这些患者,我们比较了两次采集的 MA。在我们的研究中,相当一部分患者的 MA≥1 毫米,正面投影为 80%,侧面投影为 87.9%。在接受第二次采集的患者中,第二次图像的 MA 明显降低。我们的方法通过一种简单而可靠的解决方案,可以在 EOS 图像上精确地检测 MA。我们的方法可以提高脊柱测量的可靠性,并降低因成像质量低而导致误诊的风险。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/52f7/10902286/ea0c14d610be/41598_2024_55373_Fig1_HTML.jpg

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验