Development of real-time tumor tracking algorithm for respiratory gated particle therapy

我们开发了一种基于人体数字幻象的肿瘤实时跟踪算法。由于真实的人体器官运动非常复杂，我们通过使用反映临床情况的多个数据集来改进算法。而患者的辐射暴露是x射线透视成像的另一个主要问题，因为相对较大的辐射暴露可能需要提高肿瘤跟踪计算的准确性。为了克服这一问题，我们开发了图像处理算法，以提高透视图像的质量，从而减少辐射暴露的大小，从而降低肿瘤跟踪的准确性。

We developed the realtime tumor tracking algorithm based on the human numeric phantom with intrafractional organ motions. Since real human organ motion is much complex, we improved our algorithm by using several data sets reflected to clinical situation. While radiation exposure to the patient is another major problem in X-ray fluoroscopic imaging because A relative large magnitude of radiation exposure may be necessary to improve tumor tracking calculation accuracy. To overcome this problem, we developed the image processing algorithm to improve quality of fluoroscopic image, as a result, this technique could reduce the magnitude of radiation exposure with prevent the accuracy of tumor tracking.