Classification of Multimodal Brain Images employing a novel Ridgempirical Transform

Abstract

With the evolution of technology, the assistance of hi-tech computers in the medical field occasionally involves image fusion methods. Detection and diagnosis of a disease with a single image can be tedious and difficult for doctors but with the adaptation of medical image fusion, a path for additional improvements can be paved. In this paper, the authors have proposed a Ridgempirical transform where filter banks are fused, & classified using machine learning Technique. The objective of this research is to implement different pre-processing techniques on CT-MR images of the same patient. The filter banks and spectrum are evaluated using Ridgelet Empirical Wavelet Transform (EWT) which was fused. The images are classified using Support Vector Machine. 89.5% and 86.5% of accuracy are obtained using top-hat and morphological transforms respectively. Authors have also tried other pre-processing techniques but the results employing top hat transform outperform the other techniques. To validate the proposed algorithm, the authors have used a fused CT-MR image which was pre-processed using the top-hat transform technique, and 92.1% accuracy is observed.