Least Squares Support Vector Machine Classifiers
Neural Processing Letters
Efficient computations for large least square support vector machine classifiers
Pattern Recognition Letters
Fast parametric elastic image registration
IEEE Transactions on Image Processing
Support vector machine with adaptive parameters in financial time series forecasting
IEEE Transactions on Neural Networks
Hi-index | 0.00 |
In view of the characteristics of the local nonlinear distortion about medical CT and MRI tumor image registration, the method of Least Square Support Vector Machines (LS-SVM) was used to register the images. The verge point of tumor was marked using vector machine weights in this algorithm, and the eigenvalues of tumor images were obtained. The difference of corresponding feature points between CT and MRI tumor image was eliminated adopting the least square algorithm, which not only can effectively remove geometric deformation of the images, but also can be adaptive correct the errors caused by the positioning of feature points. Finally, we had experimented on the Matlab platform, the results show that the algorithm has higher registration accuracy and can meet medical tumor registration requirements. It also play guidance role in image fusion and tumor targeted therapy, which has important application in clinical medicine.