作者: Wu Zhou , Yaoqin Xie , None
关键词: Curve fitting 、 Robustness (computer science) 、 Process (computing) 、 Image-guided radiation therapy 、 Mathematics 、 Artificial intelligence 、 Computer vision 、 Hermite polynomials 、 Boundary (topology) 、 Mathematical optimization 、 Image (mathematics) 、 Radiation treatment planning
摘要: The accurate contour delineation of the target and/or organs at risk (OAR) is essential in treatment planning for image-guided radiation therapy (IGRT). Although many automatic approaches have been proposed, few them can fulfill necessities applications terms accuracy and efficiency. Moreover, clinicians would like to analyze characteristics regions interests (ROI) adjust contours manually during IGRT. Interactive tool necessary such cases. In this work, a novel approach curve fitting interactive proposed. It allows users quickly improve by simple mouse click. Initially, region which contains interesting object selected image, then program automatically select important control points from boundary, method Hermite cubic curves used fit points. Hence, optimized be revised moving its interactively. Meanwhile, several methods are presented comparison. Finally, order delineation, process refinement based on maximum gradient magnitude All towards positions with magnitude. Experimental results show that possess superior performance proposed platform accuracy, robustness, time calculation. real medical images demonstrate efficiency, robustness clinical applications.