Clustering MRI images with principal component analysis methods

It?s problem that the images have a complicated high dimensional structure in image clustering. Because of this, the dimensions of MRI images must be reduced. The aim in the thesis is to implement PCA and image clustering methods and compare the methods. Different PCA and image clustering methods were implemented in Matlab. MRI images were used in the thesis. In the beginning, PCA methods were implemented on MRI images. The dimensions of MRI images were reduced by using PCA methods without much loss of information. After PCA methods were implemented, image clustering methods were implemented on MRI images. In the thesis five PCA methods (General Hebbian Algorithm, Adaptive Principal Component Analysis, Expectation-Maximization Principle Component Analysis, Probabilistic Principal Components Analysis and True-PCA) and two image clustering methods (K-Means and Fuzzy C-Means) were implemented. Keywords: Principal Component Analysis, Clustering, MRI