低剂量锥束CT扫描参数个体化优化算法

千伏锥束CT引导可提高患者在数十次放疗间的体位可重复性，但其累积的成像剂量导致医生不得不减少引导频次，治疗精度也随之降低。文献提示相同扫描参数下患者体型与器官剂量负相关，本组前期研究已将该关系拟合成数学模型并发表。本研究拟利用该模型进一步建立并验证基于患者体型的扫描参数个体化优化算法，从而无需硬件升级即可以较低剂量获取满意图像。主要研究内容包括：1. 利用蒙特卡洛方法模拟锥束CT关键结构及标准参数，并以测量数据进行剂量学验证。2. 利用批量患者CT数据建模计算器官剂量，并与体型及扫描参数拟合。3.利用多维共轭梯度搜索算法，基于剂量/噪声比模型和用户可调权重，计算满足临床需求的个体化优化参数。4. 使用仿真模体验证剂量、图像质量及配准精度，并以此反馈修改优化算法。该工作预期结果将揭示扫描参数影响器官剂量及图像质量的规律，其潜在应用价值是降低患者辐射风险，提高影像引导率和放疗精度。

Kilo-voltage Cone Beam CT (CBCT) improves the inter-fractional reproducibility of patient positioning during radiotherapy. However, the accumulated imaging dose and risk to normal tissue impede its frequent application hence undermine the treatment accuracy. Previous studies have reported the inverse correlations between the patient dimension and organ dose scanned by the same protocol, which have been regressed as dosimetric models by our previously published work. Based on these models, this project aims to develop and validate a patient-size specific protocol optimizer which reduces dose but maintains adequate image quality, without any hardware upgrading or imaging post-processing. The workflows include: 1. Develop and validate a Monte Carlo program which simulates the key components of a Varian CBCT and its standard protocols (done); 2. Calculate the organ dose based on a large number of patient CT data and structure sets, and generate dosimetric models as functions of body dimensions and scanning parameters (partially published); 3. Based on the dose/noise model and a user-adjustable factor, conjugated gradient searching algorithm in multi-dimensions is used to calculate the personalized optimized scanning protocol (partially done); 4. Validate the radiation dose, image quality, and impact on the registration accuracy using both adult and pediatric anthropomorphic phantoms, which provide feedback information to further improve the algorithm (Partially done). The expected results may reveal the mechanism of how scanning protocols impact the radiation dose and image quality. Its potential application may benefit cancer patients by reducing their imaging risk, and improving the treatment accuracy as a result of the increased frequency of image guidance in radiotherapy.