关键词：算法;临床医学;效率;反演;数学模型;模型;调制;优化;患者;规划;前列腺癌,前列腺;Vmat(体积调制电弧疗法)
摘 要：Inverse planning is at the heart of prostate Volumetric Modulated Arc Therapy (VMAT) treatment procedure and critically determines its level of success. As practiced now, the capacity of VMAT is greatly underutilized because of inferior computing performance of existing optimization methods. An alternative mathematical approach that improves both the efficiency and the efficacy is needed and is the center of this research. We propose to develop a new innovative inverse planning tool, based on the novel idea of superiorization, to replace the classical constrained optimization approaches employed in clinics today for prostate VMAT cases. Year 1 of the training award focused on formulating the VMAT problem as a constrained superiorization problem and on the development of a framework of fast converging inverse planning algorithms. Empty solution sets and infeasibility constraints that often exist in real-world applications were incorporated into the model. The new framework was proven mathematically and its efficacy was demonstrated when it was compared to a classical optimization method. The superiorization methodology was Implemented, tested and evaluated on a previously treated prostate case where good initial results were obtained.