Title :
3-D Reconstruction of Encephalic Tissue in MR Images Using Immune Sphere-Shaped SVMs
Author :
Guo, Lei ; Li, Ying ; Miao, Dongbo ; Zhao, Lei ; Yan, Weili ; Shen, Xueqin
Author_Institution :
Province-Minist. Joint Key Lab. of Electromagn. Field & Electr. Apparatus Reliability, Hebei Univ. of Technol., Tianjin, China
fDate :
5/1/2011 12:00:00 AM
Abstract :
In the brain MR images, the boundary of each encephalic tissue is highly irregular. Traditional 3-D reconstruction algorithms are challenged. Owing to its powerful capacity in solving nonlinearity problems, the sphere-shaped support vector machines (SSSVMs) is applied in the 3-D reconstruction. Selecting parameters for SSSVM and the kernel function, however, is a complicated issue. Appropriate parameters can make the model more flexible and help to obtain more accurate data description. In this study, immune algorithm (IA) is used in searching for the optimal parameters. Immune SSSVM (ISSSVM) is proposed to reconstruct the 3-D encephalic tissues in MR images. As shown by the experiment of this study, each encephalic tissue can be reconstructed efficiently, and satisfied accuracy and visual effect can be obtained.
Keywords :
biological tissues; biomedical MRI; image reconstruction; medical image processing; neurophysiology; support vector machines; 3-D reconstruction; MR images; encephalic tissue; immune algorithm; immune sphere-shaped SVMs; kernel function; nonlinearity problems; optimal parameters; support vector machines; visual effect; Accuracy; Image reconstruction; Immune system; Kernel; Support vector machines; Three dimensional displays; Training; 3-D reconstruction; encephalic tissue; immune algorithm (IA); support vector machine (SVM);
Journal_Title :
Magnetics, IEEE Transactions on
DOI :
10.1109/TMAG.2010.2072776
