Key Technologies Of Developing Image-guided Radiotherapy Based On Infrared Guiding And X-ray Imaging

With the development of imaging technologies, it is becoming more and more frequent to acquire different types of images during tumor treatment in radiation therapy. While it has always been a hot topic on how to take the most advantage of these images to better guide radiotherapy. And the accuracy has been improved with the ongoing improvement of image guided techniques. However, due to the fractionated treatment and body movement especially the breathing motion within one fraction, which has introduced great uncertainties to the treatment delivery and affect the accuracy. Especially, with higher dose in each fraction in stereotactic radiotherapy, it is urgent to make a better control of the motion and movement to reduce the range of margin and the random error during each treatment in order to improve coverage of the tumor target and spare more of the normal tissue. Aiming at a solution to the issues mentioned above, research has been done in this study on the positioning and tracking technologies as well as adaptive planning to improve current treatment techniques for image guided radiotherapy.Firstly, a positioning and tracking system based on infrared tracking and X-ray imaging verification was developed according to the stereotactic imaging principle and coordinate translation, proving guidance for patient positioning as well as motion correction. A software system was correspondingly built up including patient information importation, X-ray image capturing, marker-based image registration and also breathing motion management. Using our system, it can achieve fast and accurate patient setup and real-time monitoring of patient position during the treatment, offering reference for physicians and medical physicists.Secondly, a2D/3D registration method based on dual X-ray images was brought up. By applying multi-resolution strategy, it can acquire much faster and more accurate registration results with six dimensions. By involving quaternion and iterative closest point algorithm into the marker based registration process, it can solve the problem of registration with unmatched marker number detected by the system. Besides, in order to make a better control of motion caused by breathing, a non-parameter model based on Gaussian Kernel function was developed to predict the free breathing movement. Tests were done showing that the model we used can efficiently predict and adjust the changes happened during the treatment process with a precision of2mm. Thirdly, we applied the adaptive re-planning technique to clinical lung stereotactic radiotherapy, evaluating and optimizing the deformation existed in inter-fractions of each treatment. Comparisons were made between the original plan and the optimized re-plan, the results showed that the tumor coverage by prescription dose was improved by7.56%after the re-planning, and the organs at risk got more spared as well.With the opportunities and challenges in image guided radiotherapy, our positioning and tracking system based on infrared and X-ray imaging guiding was verified to be feasible and efficient, therefore can provide a better way for patient motion management with a lower cost and higher accuracy.