Multimodal medical image fusion using guided filter and curvelet transform

Abstract

The results of resonance magnetic imaging and computerized tomography of the target organ provide complementary information about this organ that helps the radiologist in the diagnosis process. Despite the information provided by these two techniques, the radiologist needs a single sensor result containing the information of the CT and MRI image for better diagnosis of the disease. Image fusion is the process of merging complementary data of several sensors into a unique image. In this study, we propose a new approach for fusing CT and MRI of brain images using a guided filter and curvelet transform. Our method is based mainly on three basic steps, which are as following: Firstly, Extracted detail layers from each input image adopting a guided filter. Secondly, based on removing the blurred images from the input images, clearer images are obtained. Finally, the images are combined using the curvelet transform. The proposed method has been compared to effective fusion methods. Through the obtained qualitatively and quantitatively results, the proposed method showed a good result compared to other methods of fusion.