Implementation of High-Dimensional Feature Map for Segmentation of MR Images

Abstract

A method that considerably reduces the computational and memory complexities associated with the generation of high dimensional (≥3) feature maps for image segmentation is described. The method is based on the K-nearest neighbor (KNN) classification and consists of two parts: preprocessing of feature space and fast KNN. This technique is implemented on a PC and applied for generating threeand four-dimensional feature maps for segmenting MR brain images of multiple sclerosis patients.

Generalized fuzzy clustering for segmentation of multi-spectral magnetic resonance images

“…Additional dimensions (such as FLAIR images) in the feature space may be required to increase the separability of more classes such as lesions [3,13,16].…”

“…The Lagrangean multiplier is adopted to include the constraints into the optimization, and the augmented objective function becomes (14) Taking the derivative of F m with respect to u ik for p >1, and equating to zero, and with the constraint we get (15) For a positive definite matrix L, for any vector x, we know that (16) Taking the derivative of F m with respect to v i and equating to zero, and using the result in Eqn.…”

“…It is possible to improve the quality of segmentation by combining information from images with multiple contrasts [3,13,16,32]. Feature map-based classification techniques for MR image segmentation have attracted considerable attention because they are fast, simple to implement, and allow expert's input in tissue classification.…”

Improving segmentation accuracy for magnetic resonance imaging using a boosted decision tree

“…The accuracy rate was quantified as the overlap fraction (Anbeek et al, 2004(Anbeek et al, , 2005He et al, 2005;Liu et al, 2005) and is defined as:…”

“…Another wavelet application has been used to design attribute vectors as spatial features of voxels for determining correspondence in 3D brain MR images (Xue et al, 2004). Segmentation applications include tissue volume quantification and 3D spatial structure reconstruction, which greatly aid in disease diagnosis (Joliot and Majoyer, 1993;Tang et al, 2000;Yoo et al, 2001;Zoroofi et al, 2001Zoroofi et al, , 2004Archibald et al, 2003;Mohr et al, 2004;Ali et al, 2005;Andrey and Maurin, 2005;He et al, 2005;Shan et al, 2005;Noulhiane et al, 2006).…”