Application of karhunen-loeve transormation for human MRI analysis
Anton Oleksandrovich Popov, Ihor Eduardovych Krashenyi, Javier Ramı́rez, Huan Manuel Gorriz · Electronics and Communications · 2014
The aim of this work is to develop new method for feature extraction from MRI images based on Karhunen-Loeve transform. Application of Karhunen-Loeve transform for multidimensional MRI images feature extraction is presented. The main result of this work is that the first basis function has the major contribution into decomposition of MRI picture, the next basis functions contributions are decreasing with their number. Recommendations for feature extraction using proposed approach for diagnosis and classification of brain diseases are given.References 11, figures 1.