Quantitative evaluation of small spherical structures in PSF-based CT image simulation

提出了一种基于点扩散函数（PSF）的三维CT图像仿真技术。我们从理想球的目标函数模拟CT图像，假设孤立性肺结节，并计算球体积与一个简单的阈值技术，其中阈值和体积之间的关系进行了分析。我们确定阈值（定义为T '），使得在模拟图像中获得的球体体积等于其目标函数（真实体积）。定量地表明，在保持物体密度不变的情况下，T'随物体球直径、图像重建核、层厚和层位而变化。我们的模拟技术和目前的结果将是有效的三维结节分析的基础研究。

A technique of three-dimensional (3D) CT image simulation based on the point spread function (PSF) was developed. We simulated CT images from the object functions of ideal spheres, assuming solitary pulmonary nodules, and calculated the sphere volumes with a simple threshold technique in which the relationship between the threshold and the volume was analyzed. We determined the threshold (defined as T') so that the sphere volume obtained in simulated image was equal to its object function (true volume). It was quantitatively shown that the T' changed with the object sphere diameter, the image reconstruction kernel, slice thickness and slice location while maintaining a constant density in the object. Our simulation technique and present results will be effective for basic research on 3D nodule analysis.