TY  - Data
T1  - Rigid registration of multimodal medical images based on mask optimization
A1  - Xiao Guoqing
DO  - 10.12072/ncdc.imp.db4086.2023
PY  - 2023
DA  - 2023-11-27
PB  - National Cryosphere Desert Data Center
AB  - In this paper, based on morphological operation, the small areas and part of the scanning table in the image are removed, the Otsu method is used to highlight the interested image parts, and Canny operator is used to extract the boundary information of the information rich areas. The pixel filling technology is used to get the mask needed for image registration, and OpenMP parallel technology is used to speed up the mask calculation process. Finally, the mask is applied to the reference image or floating image in the registration process. The experimental results show that the rigid registration method based on mask optimization can automatically remove most of the background image and scanning bed, save more than half of the time in image registration, and the image registration quality does not decline; Two groups of about 100 images can be registered in one minute. The method is successfully integrated into the treatment planning system deepplan in the form of dynamic link library. The results show that, on the basis of ensuring the accuracy of registration results, the rigid registration method of multimodal medical image based on mask optimization can significantly improve the speed of image registration, and the algorithm has high stability performance, which has a good clinical application prospect.
DB  - NCDC
UR  - http://www.ncdc.ac.cn/portal/metadata/03c82ad9-a55b-4ffb-b51a-aee12e916ec7
ER  -