Multi-modal medical image fusion using improved dual-channel PCNN

Area of Science:

• Medical Imaging
• Image Processing
• Computational Neuroscience

Background:

• Medical image fusion aims to combine complementary information from multiple sources to improve image quality and diagnostic accuracy.
• Existing fusion methods often struggle with preserving details and reducing artifacts when merging grayscale and pseudo-color images from different modalities.

Purpose of the Study:

• To propose a novel medical image fusion method that effectively combines grayscale and pseudo-color medical images.
• To enhance the detail preservation and diagnostic utility of fused medical images.

Main Methods:

• The proposed method utilizes the non-subsampled shearlet transform (NSST) for image decomposition.
• A novel improved dual-channel pulse-coupled neural network (IDPCNN) is employed for fusing high-pass sub-images.
• The Prewitt operator combined with maximum regional energy (MRE) is used for fusing low-pass sub-images.

Main Results:

• The proposed method demonstrated competitive and superior performance compared to 11 existing fusion approaches across 28 diverse medical image pairs.
• Qualitative and quantitative analyses confirmed the effectiveness of the fusion technique in preserving details and enhancing image quality.
• The method successfully combined information from grayscale and pseudo-color medical images.

Conclusions:

• The developed medical image fusion method, utilizing NSST and IDPCNN, offers a significant advancement in merging grayscale and pseudo-color medical images.
• The approach provides a robust and efficient solution for enhancing medical image analysis and diagnostic capabilities.
• This technique shows promise for improving the interpretation of medical images from various imaging modalities.