Medical image fusion using bilateral texture filtering

Many medical image fusion algorithms have been developed with the aim at enriching visual information for clinical diagnosis. However, these algorithms always suffer from texture loss, low contrast and pseudo-edges, which may result in the misdiagnosis in clinical applications. In order to solve these problems, a medical image fusion algorithm based on structural similarity detection (SSD), saliency detection and bilateral texture filter (BTF) is proposed. Our method enables the decomposition of source images into base and detail layers in the BTF scheme. For the base layers, a cluster-contrast based fusion rule using saliency detection is designed to preserve structure information. For the detail layers, we use the SSD to obtain the structural similarity part and dissimilarity part, respectively. An improved contrast-based saliency estimation method (CSE) is presented to fuse dissimilar textures; while the weighted least square optimization scheme (WLSO) is adopted to fuse similar textures. Finally, the fused image is obtained by performing reconstruction. The algorithm is compared with 17 state-of-the-art algorithms subjectively and objectively, and the experimental results have shown that the proposed method outperforms the comparative methods.

Automated summary

This paper proposes a medical image fusion algorithm based on structural similarity detection, saliency detection, and bilateral texture filter to solve the issues of texture loss, low contrast, and pseudo-edges in medical image fusion. The experimental results show that the proposed algorithm outperforms 17 other algorithms.