Multimodal Medical Image Fusion Based on Fuzzy Discrimination With Structural Patch Decomposition

Area of Science:

• Medical Imaging
• Image Processing
• Artificial Intelligence

Background:

• Multimodal medical image fusion integrates complementary information from diverse sources.
• Accurate fusion is crucial for enhancing diagnostic capabilities in clinical settings.
• Existing fusion methods face challenges in preserving image details and avoiding artifacts.

Purpose of the Study:

• To develop a novel multimodal medical image fusion method.
• To improve the accuracy and diagnostic value of fused medical images.
• To address limitations in current image fusion techniques.

Main Methods:

• Utilized structural patch decomposition (SPD) to extract salient image features.
• Developed novel incomplete and supplemental fusion maps using fuzzy logic systems.
• Implemented a weighted average approach for final fused image generation.
• Introduced a color image fusion scheme to prevent distortion and improve diagnostics.

Main Results:

• The proposed method effectively extracts salient features for fusion.
• Novel fusion maps significantly contribute to the final fused image quality.
• Experimental results show superior performance compared to state-of-the-art methods.
• The color fusion scheme successfully prevents color distortion.

Conclusions:

• The proposed SPD and fuzzy logic-based fusion method offers enhanced performance.
• This technique improves both subjective visual quality and quantitative evaluation metrics.
• The method holds promise for advancing multimodal medical image analysis and diagnosis.