A general distortion correction method for image modalities used in radiotherapy

In radiation therapy digital images taken both from the simulator and the verification system of the treatment unit, in respect to the planned and irradiated field, are essential in the patient's treatment. The correction of the distortion in these image modalities is a prerequisite for the quantification of the treatment set-up accuracy. The method proposed here, uses a well defined rectangular point grid at a specific distance from the focus between the object and the imaging chain. The authors have used graph theory model for the reconstruction of the whole grid due to the lack of information caused of the overlapping structures. The fact also that in most of the cases the irradiation fields in radiotherapy are bigger than the field of view of an image intensifier, makes necessary the acquisition of several segments of the planned field and their combination into a single image using pattern recognition techniques. The dependency of this distortion in respect to the distance of the image intensifier from the focus of the X-ray beam and to the angle of the simulator's gantry is demonstrated. The application of the proposed method has resulted into an 80% reduction of the distortion to 1.5 mm at the edges of the image intensifier. The performance of the model could be improved even more by considering both a radial and tangential distortion factor