Department of Electrical and Computer Engineering, University of Alberta, 9211, 116 St NW, Edmonton, AB, T6G 1H9, Canada.
School of Mechanical Engineering, Tianjin University, No. 135, Yaguan Road, Jinnan District, Tianjin, 300354, China.
Int J Comput Assist Radiol Surg. 2021 Jun;16(6):1027-1035. doi: 10.1007/s11548-021-02350-z. Epub 2021 Mar 29.
Low-dose-rate permanent-seed (LDR-PS) brachytherapy has shown a great potential for treating breast cancer. An implantation scheme indicating the template pose and needle trajectories is determined before the operation. However, when performing the pre-planned scheme intraoperatively, a change of the patient's posture will cause seed placements away from the desired locations. Hence, the implantation scheme should update based on the current patient's posture.
A numerical method of optimizing the implantation scheme for the LDR-PS breast brachytherapy is presented here. The proposed algorithm determines the fewest needle trajectories and template poses for delivering the seeds to the intraoperative desired locations. The clinical demand, such as the minimum distance between the chest wall and the needle, is considered in the optimization process.
The method was simulated for a given LDR-PS brachytherapy procedure to evaluate the optimal scheme as the number of the template poses changing. The optimization parameters of the needles' number and the implantation errors are used to adjust the algorithm outcome. The results show that the implantation schemes obtained by our method have a satisfactory accuracy in the cases of 2 or 3 template poses. The computation time is about 76s to 150s according to the number of the template poses from 1 to 3.
The proposed method can find the optimal implantation scheme corresponding to the current desired seed locations immediately once there is a change of patient's posture. This work can be applied to the robot-assisted LDR-PS breast brachytherapy for improving the operation accuracy and efficiency.
低剂量率永久种子（LDR-PS）近距离放射治疗在治疗乳腺癌方面显示出巨大的潜力。在手术前确定了指示模板姿势和针轨迹的植入方案。然而，在术中执行预规划方案时，患者姿势的变化会导致种子放置偏离预期位置。因此，植入方案应根据当前患者的姿势进行更新。
提出了一种用于 LDR-PS 乳腺癌近距离放射治疗的植入方案优化的数值方法。所提出的算法确定了将种子输送到术中所需位置的最少针轨迹和模板姿势。在优化过程中考虑了临床需求，例如胸壁和针之间的最小距离。
该方法模拟了给定的 LDR-PS 近距离放射治疗程序，以评估随着模板姿势数量的变化获得的最佳方案。优化参数包括针的数量和植入误差，用于调整算法结果。结果表明，在使用 2 或 3 个模板姿势的情况下，我们的方法获得的植入方案具有令人满意的准确性。根据从 1 到 3 的模板姿势数量，计算时间约为 76 秒至 150 秒。
一旦患者姿势发生变化，该方法可以立即找到与当前所需种子位置对应的最佳植入方案。这项工作可以应用于机器人辅助 LDR-PS 乳腺癌近距离放射治疗，以提高手术的准确性和效率。
