The improvement for extraction technology to target tissue of medical image

摘要

The current extracting approach to the target tissue of medical image has the problem of extracting speed. This paper presented a method of extracting the interesting area rapidly from the medical image file. Firstly, by combining run-length coding with block coding, the new coding could extract the medical image data rapidly, which not only saved the memory space and improved transmission efficiency, but also could read the required single pixel, or part of pixel data from the compressed data conveniently. Next, the author realized the extraction of interesting tissue based on combining threshold segmentation with seeded region growing, which could save the memory space and realize image segmentation rapidly and stably. Finally, The improvement mentioned above has been realized in 3D reconstructing software-MedLG which was developed by authors' institute and applied successfully in customized prostheses designing.