Multi-Modal Medical Image Fusion With Geometric Algebra Based Sparse Representation.

Area of Science:

• Medical Imaging
• Computer Vision
• Signal Processing

Background:

• Traditional multi-modal medical image fusion methods often neglect color channel correlations, leading to artifacts like color distortion and attenuation in fused images.
• Accurate fusion is crucial for enhancing pathological information and improving diagnostic accuracy in medical imaging.

Purpose of the Study:

• To develop an advanced multi-modal medical image fusion algorithm that preserves inter-channel correlations.
• To improve the quality and pathological detail of fused medical images compared to existing methods.

Main Methods:

• Proposed a novel Geometric Algebra based Sparse Representation (GA-SR) algorithm for multi-modal medical image fusion.
• Represented multi-modal images as multi-vectors to maintain channel correlation.
• Utilized Geometric Algebra Orthogonal Matching Pursuit (GAOMP) for sparse coefficient matrix acquisition and K-means clustering Singular Value Decomposition (K-GASVD) for dictionary learning and updates.

Main Results:

• The GA-SR algorithm successfully preserved color channel correlations, mitigating issues seen in traditional methods.
• Experimental results showed superior subjective and objective quality evaluations for the proposed fusion algorithm.
• Demonstrated significant improvements in the clarity and detail of fused multi-modal medical images.

Conclusions:

• The proposed GA-SR algorithm offers an effective solution for multi-modal medical image fusion.
• This method enhances diagnostic capabilities by providing richer pathological information to medical professionals.
• The algorithm represents a significant advancement over conventional fusion techniques.