Radiation therapy is a type of cancer treatment using radiation at different times defined as treatment sessions, distributed over different weeks. In each session, we have to determine and define the optimal treatment parameters for the patient. The aim of Adaptive Radiotherapy Treatment (ART) is to identify any change of initial parameters during the treatment course and modify the treatment plan for the purpose of maintaining optimal treatment objectives. In order to track the deformable image of biological organ such as the parotid gland, a 3D reconstruction is needed. 10 patients were scanned at the medical center of Oscar Lambret (Lille, France) using CT scan as imaging modality. The contours of the acquired images were extracted manually by the expert. Relaxed bi-cubic Bézier spline surface has been used in our study for the purpose of automatically reconstruction of the biological organ. Once the reconstruction is accomplished, the volume of the parotid gland at each session of treatment has been calculated for each patient. The obtained results show a decreasing of the volume of the parotid from one week to other one and a shifting of the detected center of gravity. These variations should be used to build a predictive model for adaptive robotized radiotherapy.

Adaptive Radiotherapy Treatment (ART) of cancerous organs is based on manipulating Autonomous Robotic System (ARS) used for radiation. After each treatment session, the biological organ is subject to deformation. Treatment efficiency depends highly on predicting successfully the shape of the deformable organ, being treated, prior to commencement of a new session. We begin our study by focusing on recent methods being implemented in adaptive radiotherapy treatment. Then we investigate the structural organization of treatment method from technological and biological perspectives by applying the concept of System of Systems (SoS). Moreover, we present our contribution to ART by means of a predictive model that anticipates the deformation of cancerous organ (Biological System) and issue appropriate control commands needed for adapting the ARS (Technological System). In order to do this issue, we propose in this paper a modeling evolution of a biological target to be treated or avoided during the robotized radiation sessions. For this, a bi-cubic Bézier spline surface method has been applied to reconstruct the organ in 3D dimension. The reconstruction of an organ will allows us to track the deformation of biological organ. The results show that the studied organ (Parotid) is smoothly reconstructed and its center of gravity is identified kinematicaly after each treatment session, in order to elaborate a predictive model-based approach for adaptive robotized radiotherapy.