Multimodal Medical Image Fusion Utilizing Two-scale Image Decomposition via Saliency Detection

Area of Science:

• Medical Imaging
• Computer Vision
• Biomedical Engineering

Background:

• Medical imaging modalities are crucial for diagnosing complex diseases by visualizing internal anatomy and physiology.
• Medical image fusion aims to enhance image contrast and visual impact for better computer processing and human interpretation.
• Existing fusion methods often struggle with noise magnification, feature preservation, and computational efficiency.

Purpose of the Study:

• To develop an advanced fusion framework for multimodal medical images (e.g., CT/MRI, MR-T1/MR-T2).
• To overcome limitations of classical fusion techniques, including feature loss, complex implementation, and high computational demands.
• To create a fused image that accurately retains salient features from source images while improving overall image quality.

Main Methods:

• A novel fusion framework employing a multi-scale edge-preserving filter and visual saliency detection.
• Decomposition of source images into base and detail layers using a two-scale edge-preserving filter.
• Fusion of base layers via addition and detail layers using weight maps from maximum symmetric surround saliency detection.

Main Results:

• The proposed method yields fused images with improved objective evaluation metrics compared to classical approaches.
• Resultant images exhibit enhanced global contrast and preserved edge contours.
• The fusion process effectively avoids ringing effects and artifacts, ensuring image integrity.

Conclusions:

• The developed fusion methodology offers significant improvements for medical image analysis.
• This approach enhances clinical accuracy and strengthens biomedical research capabilities.
• The framework has the potential to substantially advance medical practice and biological understanding.