Wavelet-Based Medical Image Fusion for Brain MR and Pseudo-Color PET

This paper presents a wavelet-based fusion method for brain MR and pseudo-color PET medical images. In the implementation, low-frequency components are combined using weighted averaging, while high-frequency components are processed through a 3×3 regional standard deviation fusion algorithm. Key steps involve decomposing both images using discrete wavelet transform (DWT), applying fusion rules separately to approximation and detail coefficients, and reconstructing the fused image through inverse DWT. This approach demonstrates superior performance in preserving structural details while enhancing functional information. A set of brain MR and PET medical images is provided as experimental data to validate the method's effectiveness. Notably, this fusion technique is not only applicable to brain medical images but can also be extended to other medical imaging modalities, showing broad application prospects in multi-modal medical image analysis.