DUAL ENERGY SCREENING FOR NODULAR DISEASE

这些研究的目的是开发和评估一个实用的 双能X线胸片筛查结节病的方法。 双能量技术用于减去骨骼或软组织。 通过利用不同的能量，从放射图像中提取结构 骨和软组织中X射线衰减的相关性。这项技术 使用在两种不同X射线能量下获取的图像来解算骨骼 和软组织成分。双能量成像的诊断优势在于 软组织结节的能见度通过消除 上覆肋骨的存在，结节的钙含量可能是 通过查看骨骼图像来确定。获取了双能量图像 与常规X光片同步，几乎没有增加或没有增加 患者剂量，从而使临床实用的筛查成为可能 程序。由此产生的双能量图像可以通过以下方式手动查看 放射科医生，可能在未来可能会通过自动化处理 结节检测计划，以识别这些患者以供进一步审查 有潜在的结节。 这一竞争续展申请中的项目完善了双重- 能源筛选技术是在本赠款的第1-3年中开发的。 这项建议中的具体研究包括：完善新的 散射校正算法，降噪算法的改进， 新的不对称屏幕-胶片盒的评价 互补常规图像，评价一种新的两版计算机 产生互补常规图像的射线照相方法， 详细评估双能量图像的图像质量，临床 常规/双能量混合探测技术在结核检测中的应用评价 人工神经网络在结核检测中的应用 网络以确定双能量成像是否提供了更好的特异性 用于计算机辅助检测。在这些研究领域中的每一个， 重点将放在对技术改进的详细评估上， 实施的实用性和与临床相关的诊断 进步。

The purpose of these studies is to develop and evaluate a practical screening method for nodular disease using dual-energy chest radiography. The dual-energy technique is used to subtract either bone or soft-tissue structures from a radiographic image, by utilizing the different energy dependence of X-ray attenuation in bone and soft-tissue. The technique uses images acquired at two different X-ray energies to solve for the bone and soft-tissue components. The diagnostic advantages of dual-energy are that the visibility of soft-tissue nodules is improved by eliminating the presence of overlying ribs, and the calcium content of nodules may be determined by viewing the bone image. The dual-energy images are acquired simultaneously with a conventional radiograph, with little or no increase in patient dose, thereby making possible a clinically practical screening procedure. The resulting dual-energy images may be reviewed manually by a radiologist, and possibly in the future may be processed by automated nodule detection schemes to identify for further review those patients with potential nodules. The projects in this competing continuation application refine the dual- energy screening techniques developed in years 1-3 of the present grant. The specific studies in this proposal include: refinement of a new scatter-correction algorithm, refinement of noise-reduction algorithms, evaluation of new asymmetric screen-film cassettes for acquiring the complementary conventional image, evaluation of a new two-plate computed radiography approach for producing the complementary conventional image, detailed evaluation of image quality in dual-energy images, clinical evaluation of nodule detection using the hybrid conventional/dual-energy configuration, and evaluation of nodule detection using artificial neural networks to determine if dual-energy imaging provides improved specificity for computer-aided detection. In each of these areas of research, emphasis will be placed on detailed evaluation of technical improvement, practicality of implementation, and clinically-related diagnostic improvement.