Fusion of SPECT and MRI images using integer wavelet transform in combination with curvelet transform

Abstract

Medical image fusion is used to derive useful information by fusing the images obtained from different scanners. The integer wavelet transform is shift invariant and it is a lossless transform. The fast discrete curvelet transform extracts the edge information and can preserve the details from the source images. This paper aims to apply the fast discrete curvelet transform to a multiresolution image which is obtained by applying integer wavelet transform to the source images, so that the fused image will provide all the details with clear edge information. The performance of the proposed method is compared with the existing image fusion methods. It is validated quantitatively using the metrics like entropy, standard deviation, average gradient, average, mutual information and universal image quality index. The simulation results show that the proposed method performs better than the existing methods.

Keywords:

• Integer wavelet transforms
• Fast discrete curvelet transforms
• Multimodality
• Entropy
• Magnetic resonance imaging
• SPECT
• Mutual information
• Universal image quality index