Spinks T, Guzzardi R, Bellina C R
MRC Cyclotron Unit, Hammersmith Hospital, London, UK.
Eur J Nucl Med. 1989;15(11):750-5. doi: 10.1007/BF00631770.
In positron tomographic images, the ability to differentiate closely lying structures, the spillover of activity from a region into adjacent regions and the reduction in apparent isotope concentration in small structures are all dependent on spatial resolution. Resolution in the reconstructed image is affected by (i) detector size, (ii) the spatial sampling used (e.g. stationary, wobble), (iii) the amount of smoothing in the reconstruction process (or subsequent to reconstruction) and (iv) the image pixel size. Under ideal conditions, modern commercial tomographs can produce a reconstructed spatial resolution of 5 mm or less. However, this is rarely realizable in a clinical study due to the inadequacy of counting statistics and the amplification of statistical noise. In practice, a smoother filter has to be used. This paper presents a summary of practical measurements of spatial resolution, and the related count recovery, performed on recent generation positron tomographs. It is intended to contribute to the definition of methods of measuring these parameters which is part of an on going concerted action in positron tomography supported by the European Commission.
在正电子断层图像中,区分紧邻结构的能力、活性从一个区域扩散到相邻区域以及小结构中表观同位素浓度的降低都取决于空间分辨率。重建图像中的分辨率受以下因素影响:(i)探测器尺寸;(ii)所使用的空间采样(例如静止、摆动);(iii)重建过程中(或重建后)的平滑量;以及(iv)图像像素大小。在理想条件下,现代商用断层扫描仪能够产生5毫米或更小的重建空间分辨率。然而,由于计数统计的不足和统计噪声的放大,这在临床研究中很少能够实现。实际上,必须使用更平滑的滤波器。本文总结了对新一代正电子断层扫描仪进行的空间分辨率及相关计数恢复的实际测量。其目的是为测量这些参数的方法的定义做出贡献,这是由欧盟委员会支持的正电子断层成像中正在进行的协同行动的一部分。