Correcting Distorted Image And Locating Intramedullary Nail Of Robot Assisted Surgery

The research work of getting accurate information from image and the position of hole will provide theoretical and practical experiences, and will be helpful in the field of orthopaedics especially. The thesis is part of a state-aided 863 project, for improving the method of correcting distorted X-ray image and Locking intramedullary nailThis thesis introduces how to get the accurate length of two points after analysis the deforming of X-ray image. Firstly, extract and sort circles of calibration disc's X-ray image by the PHT method. Secondly, this thesis purposes a new deformation correction method basing on neural net. It can get the mapping between spatial points and points in the image using this method after getting input, output and condition without considering other factors. BP net is a particular type of neural network model, known as a "feed-forward back-propagation network". A BP network learns by example, that is, we provide a learning set a big group of circles that consists of some input examples and the known-correct output for each case. So, we use these input-output examples to show the network what type of behavior is expected, and the BP algorithm allows the network to adapt. We get the accurate result through BP net At last.We count spatial relationship using simulating the track between transmitting terminal and sink by curved line. It needs complex instruments to achieve more planes. So researchers always fix registration attachment on manipulator of system. Then registration attachment can move from transmitting terminal and sink of C-arm because of flexibility of manipulator. We mount marker cube on the guiding robot and calculate the position and orientation of hole in intramedullary nail.Theoretically,we can calculate the transfer matrix between Cube frame and Robot-Tool frame, because the relative position between marker cube and guiding robot is fixed. But error of mechanical interface is too serious to accomplish locking hole. We solve this problem by using neural net. We transfer frame...