Multimodality Medical Image Processing Platform

The development of medical imaging facilities and techniques of image processing has been enforcing the advancement of medical diagnose and therapy. Nowadays, assisted by multi-modeled medical images (CT, MRI, etc.), doctors can examine pathological changes from different visual angles and even in 3D mode.Various kinds of imaging equipments provide different information about the same viscera, which helps doctors judge with more accuracy. And the obligation and challenge of image processing is to find out efficient method to process those images before they are available and accessible to doctors. The methods for medical image processing (MIP) are now swarming. According to different purposes it serves, MTP technique can be roughly divided into three part—segmentation, registration, data association, and 3D visualization.To fully take advantages of the existing methods and at the same time to avoid repeated work load, some research institutions have begun to integrate valuable algorithms into a uniform software platform. Successful examples are ITK and VTK. However, both ITK and VTK have their drawbacks respectively. ITK is a tool kit with no friendly interface, while VTK a general visualization tool kit not exclusively for MIRIn light of this, we set our purpose to develop a multi-model MIP platform. To take the action, we combine the virtues of VTK and ITK and construct above them a practical and open integrated MIP platform. So that efficient cooperation and communication can be practiced upon it.In this thesis, we focus on the framework design of the multi-model MIP platform, in particular, a MIP software called 3DTools is proposed .This thesis contains seven chapters. Chapter two and three are introduction to ITK and VTK, and some image processing examples are shown. Chapter five and six discuss the overall framework and software development of the multi-model MTP platform. Here are more details: Chapter one have a simple introduction of MIP and the work we have done. Chapter two gives profound discussion about ITK and some new and impressive algorithms of segmentation and registration are shown by using ITK. Chapter three discusses VTK in details and we focus on its 3D visualization ability. Chapter four analyses the disadvantages of VTK and ITK, which is also the problem our work tries to solve. Chapter five proposes the framework of the multi-model MIP platform. The realization concerns and solutions to the key problems are discussed. In Chapter six, the development and design of MIP 3DTools is proposed. A conclusion is contained in the last chapter.