Multi-modal Medical Image Fusion Approach Utilizing Gradient Domain Guided Image Filtering

Area of Science:

• Medical imaging
• Image processing
• Computer vision

Background:

• Current multimodal medical image fusion methods often neglect color preservation, focusing primarily on edge details.
• This limitation can reduce the diagnostic value of fused medical images.

Purpose of the Study:

• To develop a multimodal medical image fusion algorithm that enhances color preservation.
• To improve the overall fidelity and spectral integrity of fused medical images.

Main Methods:

• A multi-channel fusion algorithm utilizing gradient domain-guided image filtering.
• Image decomposition using wavelet transforms and layer construction.
• Application of various fusion rules followed by inverse wavelet transformation.

Main Results:

• The proposed algorithm achieved superior performance across multiple objective metrics (MSE, CCI, PSNR, SSIM, DD, SM).
• Consistently outperformed existing alternative fusion methods in quantitative evaluations.

Conclusions:

• Experimental results confirm the algorithm's effectiveness in preserving edge details.
• The fusion algorithm successfully maintains image fidelity and spectral integrity, enhancing diagnostic potential.