Study On Key Techniques Of Multi Leaf Collimator Radiotherapy System

3D conformal radiotherapy (3DCRT) and intensity modulated radiotherapy (IMRT) are important portion of modern precision radiotherapy, they are effective measures to improve radiotherapeutic gain.3DCRT and IMRT are carried out mainly by multi-leaf collimator (MLC). In this dissertation, key techniques of MLC conformal and intensity modulated radiotherapy system are discussed, including the positioning of patients and diagnostic image magnification, cross-sectional interpolation for multi contour slices and contour simplification, gray-level image interpolation in consecutive image slices, 3D data visualization, beam arrangement and optimization of MLC leaf end shape, etc.. The system design of ModernTPS treatment planning system (TPS), the main contents in MDMLC multi-leaf collimator mechanical design and the development of the control system are also introduced in this dissertation.CT/MRI diagnostic images are transferred to TPS, they are fundamental data set for TPS. Pre-process should be carried out for these images first, which is mainly focused on the detection of positioning marks left by positioning facility. Two methods for the detection of positioning marks in diagnostic images, an auto one and a semi-auto one, are presented. By these methods, the localization of image slices can be realized, and the status of the positioning facility can be monitored. For convenient observation of 2D diagnostic images, a 2D image magnification algorithm based on fairing and continuous restoration of edge curves is presented. As a result of this approach, the edges of the magnified image keep sharp and fairness, which is better than that in conventional methods.Multi-contour process is an important task in TPS. In order to represent the 3D structure with higher precision, it's often necessary to construct some cross-sectional contours. Linear interpolation between contours causes non-smooth morphing at original slices. A fairing cross-sectional interpolation approach for multi contour slices is presented, by which faired morphing contours can be obtained. With respect to the data redundant in contours, a feature reserving simplification method is presented, by which contours are approximated by segmented lines while preserving precision. Surface is reconstructed with adjacent contours synchronous marching algorithm based on the simplified contours.3D regular volume data is the fundamental data set for TPS. In order to eliminate the disadvantage of edge blurring by traditional gray-level image interpolation in the construction of regular volume data, a matching approach for interpolation in consecutive image slices is proposed based on image histogram, better results can be obtained than traditional algorithms. Marching cubes (MC) and marching tetrahedral (MT) algorithm are used to extract...