成像体模
矢状面
像素
图像质量
冠状面
标准差
磁共振成像
翻转角度
物理
光学
放大倍数
信号(编程语言)
图像分辨率
核磁共振
数学
计算机科学
人工智能
图像(数学)
医学
解剖
统计
放射科
程序设计语言
作者
Barrie Condon,James C. Patterson,D. Wyper,A. Jenkins,Donald M. Hadley
出处
期刊:British Journal of Radiology
[British Institute of Radiology]
日期:1987-01-01
卷期号:60 (709): 83-87
被引量:119
标识
DOI:10.1259/0007-1285-60-709-83
摘要
Image non-uniformity in magnetic resonance imaging (MRI) is a recognised problem when using surface coils but is rarely mentioned with respect to standard head coils, yet it can be a source of error in clinical interpretation of images. Figure 1 demonstrates the type of artefacts due to image non-uniformity for a large, homogeneous solution of water and copper sulphate. The image was acquired in the coronal section using a standard head coil, and vertical (i.e. read direction) and horizontal (i.e. phase encode direction) pixel intensity profiles through the centre of the image are illustrated. The standard deviation of pixel intensities, expressed as a percentage of the mean value, is greater than ±27% along the 23 cm length of this phantom. Figure 2 demonstrates how these image nonuniformities can affect the sagittal image of a normal volunteer. The display window width and level have been chosen to produce optimum visual contrast at the level of the cerebellum and brain stem but, as a result to image non-uniformity, this produces pronounced loss of signal in the signal in the neck and the vault of the skull. Quantitative and even qualitative comparison of pixel intensities across the image in the vertical direction (along the spine for example) are certainly invalid without some form of non-uniformity correction. Image quality assurance measurements, for example using phantoms to determine plane shape and thickness, will also be affected by this problem.
科研通智能强力驱动
Strongly Powered by AbleSci AI