Research On Key Technologies For Optical-CT Dual-modality Imaging Processing

Multimodality imaging has become an important part of modern medical imaging technology.A non-invasive method can be used in the in vivo small animal optical imaging system to track the fluorescent markers for a long time.However,it can not provide structural information.The combination of CT and optical images can improve the accuracy of the optical reconstruction and provide the observation of the tissue from both structural and functional perspectives.The main works of optical-CT dual-modality imaging data processing in this paper are as follows:(1)Developed a new mouse fixation device for optical-CT dual-modality imaging.By using the markers in mouse fixation device,it realized the unification of the coordinates between two systems.(2)Researched the preprocessing methods of optical and CT images.In order to remove the fluorescence noise,the mean square error ratio and Criminisi algorithm were used to extract the target fluorescence image.In CT preprocessing,the mouse fixation device was removed by the excessive segmentation factor,and the surface of mouse was restored by the morphological method.(3)Proposed a depth-based permissible region selection method.It can predict the source depth according to the linear relationship between the results of diffusion approximation equation by Green function method and the fluorescence intensity collected by CCD camera.And the depth was set into CT to mark permissible region on the dual-modality coordinate relationship.It improved the speed and accuracy of setting permissible region.(4)Studied the algorithm of inner source reconstruction in AOIS.Due to the ill posed problem of reconstruction,the source was carried out by using the permissible region and the StF-Lasso algorithm based on l1 regularization.(5)The reconstruction results were unable to indicate the distribution and changes of the source because they were discrete points.We proposed an optical-CT data fusion method based on source tomography images and realized the volume rendering using GPU programming.The mouse experimental results showed that the optical-CT dual-modality imaging processing algorithm in this paper can realize the accurate positioning and fusion display of the source.