# YUV-based SVD-VGG hybrid fusion for multimodal MRI-PET image integration

## TL;DR

This paper introduces a new method for combining MRI and PET brain images using a hybrid approach that improves detail and color while handling noise.

## Contribution

The novel SVD–VGG hybrid fusion framework integrates luminance decomposition and feature enhancement for noise-aware multimodal image fusion.

## Key findings

- The proposed method achieves high structural fidelity and perceptual quality in fused images.
- It maintains sub-second runtime while preserving color and contrast under synthetic noise conditions.
- Quantitative metrics like PSNR, SSIM, and LPIPS show consistent performance improvements.

## Abstract

Multimodal medical image fusion enhances diagnostic interpretation by integrating anatomical and functional information into a single image. This work proposes an efficient hybrid framework, termed SVD–VGG Hybrid Fusion, unifying Singular Value Decomposition (SVD) for luminance decomposition and a lightweight VGG-based feature extractor for high-frequency enhancement. Synthetic Gaussian noise (σ2=0.25) is added to MRI and Poisson noise to PET images to simulate representative acquisition degradations, while the SVD and VGG-based feature paths strengthen structural detail and functional contrast. Experiments were conducted on a single public brain dataset with image pairs resized to 256×256 for fusion and 224×224 for feature extraction. Quantitative evaluation using PSNR, SSIM, CC, and perceptual LPIPS indicates that the proposed method achieves consistent structural fidelity, perceptual quality, and color preservation while maintaining sub-second runtime per case. Although evaluated only on brain MRI–PET data and under synthetic noise conditions, the results suggest that the SVD–VGG hybrid design provides a noise-aware and color-preserving fusion strategy suitable for practical multimodal image fusion workflows.