Medical image fusion using enhanced cross-visual cortex model based on artificial selection and impulse-coupled...

Area of Science:

• Medical Imaging
• Image Processing
• Computational Pathology

Background:

• Traditional Image Composite Method (ICM) has limitations in accuracy and cost due to parameter settings.
• Complex edges and details in medical images necessitate advanced fusion algorithms for accurate pathological diagnosis.

Purpose of the Study:

• To develop an improved image fusion algorithm for enhanced diagnosis of cerebral infarcts and acute strokes.
• To address the limitations of traditional ICM in medical image analysis.

Main Methods:

• A novel image fusion algorithm combining Nonsubsampled Contourlet Transform (NSCT) and an improved Image Composite Method (ICM) was developed.
• Proposed low-frequency and high-frequency sub-band fusion rules.
• Applied the algorithm to CT/MRI image fusion and compared it with NSCT-SF-PCNN, NSCT-SR-PCNN, and Adaptive-PCNN.

Main Results:

• The proposed NSCT and improved ICM fusion algorithm demonstrated superior performance over other methods based on objective and subjective evaluations.
• The algorithm successfully fused CT/MRI images of healthy brain tissue, cerebral infarcts, and acute strokes.
• Experimental findings validated the effectiveness of the combined NSCT and improved ICM approach.

Conclusions:

• Medical image fusion using the proposed algorithm, particularly with Adaptive-PCNN, yields satisfactory results.
• The fusion technique significantly improves image clarity, edge information, contrast, and brightness.
• This approach holds good clinical value for the pathological diagnosis of complex neurological conditions.