The nature of statistical learning theory
The nature of statistical learning theory
Nonlinear component analysis as a kernel eigenvalue problem
Neural Computation
Everything old is new again: a fresh look at historical approaches in machine learning
Everything old is new again: a fresh look at historical approaches in machine learning
Brain tumor classification based on long echo proton MRS signals
Artificial Intelligence in Medicine
Hi-index | 0.00 |
This paper discusses the use of nonlinear PCA (K PCA) and regularized least squares classification (RLSC) algorithm to differentiate malignant(cancerous) from benign (noncancerous) soft tissues tumors in MRI images. In this study, we address the effect of bias fields on the PCA analysis and propose to carry out PCA in the Fourier domain so that the principal components are extracted form certain frequency coefficients of the MRI image which enable us to select certain frequency components to be used for extracting the PCA principal components. Thus, the PCA analysis in the Fourier space serves two purposes. First, it transforms the tumor subimages into a principal component space to achieve a dimensionality reduction. Second, it maximizes the separation between the malignant and the benign tumors in the feature space by removing the low frequency components of the tumor subimages that mostly contain the field inhomogeneity signal. The subimages that are projected into the principal component space are used as a training set for a regularized least squares classifier (RLSC). The RLSC classifier is evaluated by an independent tumor subimages.