Suppr超能文献

利用衰减系数图校正单光子发射计算机断层扫描中的衰减:一种新方法。

Correction of attenuation in SPECT with an attenuation coefficient map: a new method.

作者信息

Bourguignon M H, Berrah H, Berdriem B, Riddell C, Valette H, Wartski M, De Dreuille O, Jouve B, Syrota A

机构信息

Service Hôspitalier F. Joliot, Département de Recherche en Imagerie, Pharmacologie et Physiologie, Orsay, France.

出版信息

J Nucl Biol Med (1991). 1993 Mar;37(1):26-32.

PMID:8329474
Abstract

Attenuation coefficient (mu) maps, measured from transmission scan, are now becoming available. A simple method of attenuation correction is needed for routine implementation, however. Significant attenuation compensation can be mathematically obtained by dividing each actual pixel value of emission projections by the average of all the attenuation factors [exp (-sigma mu)] of all voxels along the same projection ray. This simple method, compatible with filtered back projection algorithms, was tested on simulations of cardiac and cerebral transaxial images on a Vax computer using the RECLBL library. In the models, the different structures received different activity and mu values. Three types of emission projections were generated: the ideal projections obtained by summation of the activity along each projection ray, the corresponding attenuated projections, and the projections corrected for attenuation. Comparison of projections on a pixel by pixel basis showed differences of less than 20% between the corrected and ideal projections. After reconstruction, both absolute and relative quantification were greatly improved by the correction of attenuation. Further validation of the method is in progress with actual patient data.

摘要

通过透射扫描测量得到的衰减系数(μ)图现已可用。然而,常规应用需要一种简单的衰减校正方法。通过将发射投影的每个实际像素值除以沿同一条投影射线的所有体素的所有衰减因子[exp(-σμ)]的平均值,可以在数学上获得显著的衰减补偿。这种与滤波反投影算法兼容的简单方法,在使用RECLBL库的Vax计算机上对心脏和脑断层图像模拟进行了测试。在模型中,不同结构具有不同的活性和μ值。生成了三种类型的发射投影:通过沿每条投影射线求和活性得到的理想投影、相应的衰减投影以及经衰减校正的投影。逐像素比较投影结果显示,校正后的投影与理想投影之间的差异小于20%。重建后,通过衰减校正,绝对和相对定量都得到了极大改善。该方法正在使用实际患者数据进行进一步验证。

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