Research On External Fiducially Markers-based Registration In Cardiac Interventional Surgery Navigation

Heart Intervention is an new medical technology for diagnostic and treatment of cardiovascular diseases,it made outstanding contributions to the fighting against coronary heart disease,rheumatic heart disease and other human killers,at the same time,because of surgical trauma,shorter operative time,high security,lower cost and other reasons,it has been widely used in clinical diagnosis and treatment of related diseases,which is of great clinical significance.This research is to study and realize the registration method of surgical navigation in the heart of the intervention,the main content is about external markers based image-guided surgical navigation registration,including three-dimensional reconstruction of external markers and 3D external markers registration and other key research questions.Specific works and results are as follows.(1)A method for matching external markers in cardiac angiograms is proposed after studying the transformation relations between different angle of cardiac angiography projection image.Constructing a two-dimensional characteristics descriptors for the external markers in every image,then matching external markers by using the Euclidean distance of feature descriptors in the pictures of cardiac imaging at the adjacent angle projection,the experimental results show that this method can be completed quickly and match the correct all external markers of all the different projection angles cardiac images.Furthermore,we compare the triangle chain code matching method,experiments show that the proposed method have advantage on the correct matching rate.(2)Implementing three-dimensional reconstruction of the external markers using multi-angles cardiac images.Firstly we reconstruct the coordinate positions of the markers in the C-arm three-dimensional coordinate system by using C-arm geometry calibration and least squares estimating.After matching external markers in the multi-angle cardiac images,calculating the marker coordinates according to the geometric intersection and least squares estimation theory,and then analyzing the amount of displacement caused by cardiopulmonary exercise by using the external markers distant.(3)Further studying of the ICP algorithm and proposing to remove the large displacement of external markers to improve the accuracy of registration methods.By analyzing the experimental data suggests that the large displacement amount of marker in the marked point will decrease the accuracy of registration,which does not meet the clinical needs of the phenomenon.Taking out a large amount of displacement the marker,through experiments found that this method can solve the problem of declining registration accuracy,the experimental data show that the registration accuracy can also be achieved clinical needs.(4)3D visualization software is designed and developed.Using medical imaging developer toolkit ITK and VTK to develop the regional organizations of the heart integration segmentation and registration modules,which builds a virtual three-dimensional space,and use Qt software platform designed interface.During the surgery,input the coordinates which in the position of the surgical instrument in three-dimensional space to the software platform,computer rendering three-dimensional display of the human body in space with cone represents the relative position to surgical instruments,which provides navigation information for doctors during surgery operation.