Technology Research On Compression, Retrieval And Display Of CT Data Oriented To Databases

The medical imaging is an important tool for assisting the doctor to clinical diagnose. Tomographic scanning data is the main component of medical imaging, but its amount is too huge. To construct a database is a useful way to organize and apply it. The key technology of building tomographic scanning database is compression, retrieval and display. This thesis has done in-depth studies for this. The main content includes:(1) We have put forward the plan of DICOM conversion,studied the methods of the DICOM and BMP conversion and the extraction of the necessary data element information. We have used the improved run-length encoding method and difference image compression method to compress the converted BMP, so the compression ratio has been improved.(2) We have studied the retrieval methods of the 3D geometry surface mode of this database, and put foward two retrieval methods, one based on anatomy parameter, and the other based on section multi-line contour. For the retrieval method based on section multi-line contour, we have studied two kinds of fourier descriptors, and designed the shape similarity identifying method for section multi-line contour.(3) This thesis has done studies for the displaying method of the 2D image and 3D model in MITK. The 2D image display model class and the 3D model rendering model class have been rebuilt, and the display function of MITK is enhanced. Based on this the position corresponding mechanism between three views (the data channel) has been built, realizing the associated display of the 2D image and 3D model.Using these technologies above, we have built a database management system for tomographic scanning data. Preliminary tests show that this system is stable, and the interactive operation satisfies the real-time requirement.