Suppr超能文献

细肺纹理检测中的高频边缘增强。存储磷光体数字图像与传统胸部X线摄影的对比。

High frequency edge enhancement in the detection of fine pulmonary lines. Parity between storage phosphor digital images and conventional chest radiography.

作者信息

Oestmann J W, Greene R, Rubens J R, Pile-Spellman E, Hall D, Robertson C, Llewellyn H J, McCarthy K A, Potsaid M, White G

机构信息

Department of Radiology, Massachusetts General Hospital, Harvard Medical School, Boston.

出版信息

Invest Radiol. 1989 Sep;24(9):643-6. doi: 10.1097/00004424-198909000-00001.

Abstract

Fine linear structures represent a severe test of the minimum spatial resolution that is needed for digital chest imaging. We studied the comparative observer performance of storage phosphor digital imaging (1760 X 2140 pixel matrix, 10 bits deep), and conventional radiography (Lanex medium screen, Ortho C film) in the detection of simulated fine pulmonary lines superimposed on the normal chest when exposure factors were identical (20mR skin entrance dose at 141 kVp). Receiver operating characteristics analysis of 2160 observations by six readers found that high frequency edge-enhanced digital images (ROC area: 0.78 +/- 0.06) performed better than unenhanced digital images (ROC area: 0.70 +/- 0.07) (P less than 0.01 for paired t-test), and that edge enhanced digital images performed on a par with conventional radiography (ROC area: 0.78 +/- 0.09). We conclude that for the detection of fine linear structures, storage phosphor digital images can perform on a par with higher resolution conventional chest radiographs when a high frequency edge-enhancement algorithm is employed.

摘要

精细线性结构对数字胸部成像所需的最小空间分辨率构成了严峻考验。我们研究了存储磷光体数字成像(1760×2140像素矩阵,10位深度)和传统放射摄影(Lanex中速屏,柯达C型胶片)在检测叠加于正常胸部的模拟精细肺纹理时的观察者比较性能,此时曝光因素相同(141 kVp时皮肤入口剂量为20mR)。六位读者对2160次观察结果进行的受试者工作特征分析发现,高频边缘增强数字图像(ROC曲线下面积:0.78±0.06)的表现优于未增强数字图像(ROC曲线下面积:0.70±0.07)(配对t检验P<0.01),且边缘增强数字图像的表现与传统放射摄影相当(ROC曲线下面积:0.78±0.09)。我们得出结论,对于精细线性结构的检测,当采用高频边缘增强算法时,存储磷光体数字图像的表现可与更高分辨率的传统胸部X线片相媲美。

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