Fast image registration based on particle swarm optimization algorithm and its application
Image registration is critical for clinical medical diagnosis and assisting surgical procedure.Using simulated X-ray and digital subtraction angiography(DSA)to generate digitally reconstructed radiograph(DRR)would improve the accuracy and speed of medical image registration.Herein an approach of using shading language to simultaneously generate the 3D render image and DRR from simulated X-ray or DSA is presented.The complex optimization and registration calculations are completed using the parallel calculation language(OpenCL),and the image processing unit(GPU)is used for DRR generation and 2D/3D registration.Experimental results show that the proposed method which relies on simulated X-ray or DSA for image registration can provide rapid and accurate visual guiding for surgical procedure.
NETL
NSTL
万方数据
