Correlation Coefficient Measure of Mono and Multimodal Brain Image Registration using Fast Walsh Hadamard Transform

Abstract

A bundle of image registration procedures have been built up with enormous implication for data analysis in medicine, astrophotography, satellite imaging and little other areas. An approach to the problem of mono and multimodality medical image registration is proposed, with a fundamental concept Correlation Coefficient, as a matching measure. It measures the statistical dependence or information redundancy between the image intensities of corresponding voxels in both images. Maximization of CC is a very broad and dominant norm. As no assumptions are made regarding the nature of this dependence and no limiting constraints are imposed on the image content of the modalities involved. The accuracy of the CC criterion is validated for rigid body registration of computed tomography (CT), and magnetic resonance (MR T1 and T2) images by comparison with the registration solution. Experimental results prove that subvoxel accuracy with the reference solution can be achieved completely automatically without any preprocessing steps that make this process ensemble for medical applications.

Multi-Class Diagnosis of Skin Lesions Using the Fourier Spectral Information of Images on Additive Color Model by Artificial Neural Network

“…Next, all the particular values of µ F M (i) are concatenated to create a particular features vector for each class, such as (16) ). Once the V µ F M of all the classes have been calculated, each particular sample of each class (FV M ,l ) must be correlated with all the V µ F M to calculate the Pearson's correlation coefficient, as shown (6) [34], [35]:…”

Evaluation of Optimal Control-Based Deformable Registration Model

“…As the name suggests, the CC quantifies the similarity between the template image and the reference image based on how well they correlate with each other. It measures the statistical dependence or information redundancy between the image intensities of corresponding pixels in both images [14]. This similarity measure is computed as follows:…”

“…This similarity measure is very appropriate in multimodal applications. No limiting constraints are imposed on the image content of the modalities involved [14].…”

Fast computation of residual complexity image similarity metric using low‐complexity transforms

“…Furthermore, small block sizes lend themselves well for approximate multiplier‐free transforms for which fast hardware implementations are available [11–15]. In biomedical applications, multiplierless transformations have been considered in constructing phase and amplitude images used to evaluate abnormalities of cardiac contraction [16], pulmonary sound analysis [17], DNA sequencing [18] and brain image registration [19]. Although these studies demonstrate the potential of using block processing and approximate transforms to accelerate DCT calculation, its use in computing residual complexity has not been studied before.…”