Bonfring International Journal of Advances in Image Processing

Online ISSN: 2277-503X | Print ISSN: 2250-1053 | Frequency: 4 Issues/Year

Impact Factor: 0.245 | International Scientific Indexing(ISI) calculate based on International Citation Report(ICR)

MRI Image Registration Using Fast Independent Component Analysis

K. Aishwarya and S.P. Sivagnana Subramanian

---

Abstract:

Image Registration is the process of transforming different sets of data into one coordinate system. Inter-subject multichannel image registration method is used which combines information from different modalities to produce a unified joint registration. The problem identified in the existing method is that it uses Discrete Wavelet transform and combines information only at the intensity level, whereas the proposed method uses Gabor wavelet transform and combines information at the feature level which is more robust and accurate. Four MRI (Magnetic Resonance Imaging) images T1, pd, T2 and FA are taken as input images. The first step is feature extraction using Gabor filters. The next step is the Independent Component Analysis (ICA). Fast ICA using fixed-point algorithm is used to compute the independent components. The features to be selected for fusion are determined from the ?Choose-Max? fusion by collecting the mutual information of the features present in images after ICA. This can be obtained from histogram. After Choose-Max fusion, the mutual fausse goyard information from the resultant four images are fused to a single image. This fused image has higher information when compared to the individual images which is useful for Image Registration.

Keywords: Image Registration, Gabor Wavelet, Fast ICA, Mutual Information

Volume: 2 | Issue: Special Issue on Communication Technology Interventions for Rural and Social Development

Pages: 17-20

Issue Date: February , 2012