Font Size: 

    In clinical application, radiotherapy is the mainstay of cancer treatment. The deformable image registration (DIR) is a medical imaging device which uses to assess how the tumor changes in size and location. DIR is a very important medical guide in radiotherapy treatment. The satisfactory treatment could be achieved if the exact location and extent of target region are accurately determined, thus the surrounding organs could suffer less impact.
    In computational field, the DIR has been studied over 30 years in the discipline of biomedical science. Due to the complexity of the physical phenomenon of a human body, the development of deformable image models could not be derived uniquely. Many applications did not accomplish the clinically satisfactory level. To improve the treatment flexibility, a high level of accuracy is indispensable in delivering right medical dose to each radiotherapy treatment. Thus the deformable image registration remains to be attractive and important research topics in the clinical field.
    In general, the DIR is an optimization method used to measure how the tumor changes in size and location. Several forms of deformable image registration models have been established and solved by different numerical methods. This paper adopts the meshless algorithm using the kernel-based collocation method to solve the concerned model. The application and its formulation to the classical model of DIR algorithms are outlined. The application to real-life case study and the corresponding results will be presented and discussed in our further report.

Deformable image registration, Meshless, Kernel-based collocation