谷歌浏览器插件
订阅小程序
在清言上使用

Optimizing collimator resolution/sensitivity in SPECT iterative reconstruction

Nuclear Science Symposium and Medical Imaging Conference(2013)

引用 0|浏览7
暂无评分
摘要
We present a simulations study of the collimator sensitivity/resolution compromise as it affects the image quality of hot lesions in SPECT iterative reconstruction. We investigate three parallel-beam hexagonal-hole collimators with resolution in the center of the field of view equal to 11 mm, 15 mm, and 19 mm. We scan the Esser hot-rod phantom with 60 views and we reconstruct the images using MLEM iterative reconstruction with no regularization or filtering. As the image-quality figure of merit we use the reconstructed-image contrast vs noise. We repeat the study for three levels of object-contrast (2.3, 4, and 9). We conclude that the higher-sensitivity collimator (15-mm in resolution) offers an improvement (lower noise for same contrast or higher contrast for same noise) in the reconstruction of medium-sized lesions (12-mm and 16-mm), at medium to low contrasts (4, 2.3). At high contrast (9) and low lesions sizes (8-mm) the higher resolution (11-mm) collimator is still favorable. The conclusions are task-dependent, and apply to hot lesions of the particular sizes and contrasts. Similar studies can be performed to address a range of realistic clinical problems. We intend to repeat this study for cold lesions in a cardiac anthropomorphic phantom.
更多
查看译文
关键词
cardiology,collimators,diseases,image denoising,image reconstruction,iterative methods,medical image processing,phantoms,single photon emission computed tomography,esser hot-rod phantom,mlem iterative reconstruction,spect iterative reconstruction,cardiac anthropomorphic phantom,high-sensitivity collimator,hot lesions,image-quality figure of merit,noise,object-contrast levels,optimizing collimator resolution-sensitivity,parallel-beam hexagonal-hole collimators,reconstructed-image contrast,gate,monte carlo simulations,spect,iterative image reconstruction,tomographic imaging,image resolution
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要