当前位置: X-MOL 学术J. Mech. Med. Biol. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
OPTIMIZED DUAL-TREE COMPLEX WAVELET TRANSFORM AND FUZZY ENTROPY FOR MULTI-MODAL MEDICAL IMAGE FUSION: A HYBRID META-HEURISTIC CONCEPT
Journal of Mechanics in Medicine and Biology ( IF 0.8 ) Pub Date : 2021-03-18 , DOI: 10.1142/s021951942150024x
N. NAGARAJA KUMAR 1 , T. JAYACHANDRA PRASAD 2 , K. SATYA PRASAD 3
Affiliation  

In recent times, multi-modal medical image fusion has emerged as an important medical application tool. An important goal is to fuse the multi-modal medical images from diverse imaging modalities into a single fused image. The physicians broadly utilize this for precise identification and treatment of diseases. This medical image fusion approach will help the physician perform the combined diagnosis, interventional treatment, pre-operative planning, and intra-operative guidance in various medical applications by developing the corresponding information from clinical images through different modalities. In this paper, a novel multi-modal medical image fusion method is adopted using the intelligent method. Initially, the images from two different modalities are applied with optimized Dual-Tree Complex Wavelet Transform (DT-CWT) for splitting the images into high-frequency subbands and low-frequency subbands. As an improvement to the conventional DT-CWT, the filter coefficients are optimized by the hybrid meta-heuristic algorithm named as Hybrid Beetle and Salp Swarm Optimization (HBSSO) by merging the Salp Swarm Algorithm (SSA), and Beetle Swarm Optimization (BSO). Moreover, the fusion of the source images’ high-frequency subbands was done by the optimized type-2 Fuzzy Entropy. The upper and lower membership limits are optimized by the same hybrid HBSSO. The optimized type-2 fuzzy Entropy automatically selects high-frequency coefficients. Also, the fusion of the low-frequency sub-images is performed by the Averaging approach. Further, the inverse optimized DT-CWT on the fused image sets helps to obtain the final fused medical image. The main objective of the optimized DT-CWT and optimized type-2 fuzzy Entropy is to maximize the SSIM. The experimental results confirm that the developed approach outperforms the existing fusion algorithms in diverse performance measures.

中文翻译:

多模态医学图像融合的优化双树复小波变换和模糊熵:混合元启发式概念

近年来,多模态医学图像融合已成为一种重要的医学应用工具。一个重要的目标是将来自不同成像模态的多模态医学图像融合成单个融合图像。医生广泛利用它来精确识别和治疗疾病。这种医学图像融合方法将通过不同模态从临床图像中提取相应信息,帮助医生在各种医学应用中进行综合诊断、介入治疗、术前计划和术中指导。在本文中,采用智能方法采用了一种新颖的多模态医学图像融合方法。原来,来自两种不同模态的图像应用优化的双树复小波变换 (DT-CWT) 将图像分成高频子带和低频子带。作为对传统DT-CWT的改进,通过合并Salp Swarm Algorithm (SSA)和Beetle Swarm Optimization (BSO)的混合元启发式算法称为Hybrid Beetle and Salp Swarm Optimization (HBSSO)优化滤波器系数. 此外,源图像的高频子带的融合是通过优化的2型模糊熵来完成的。成员上限和下限由相同的混合 HBSSO 优化。优化的 2 型模糊熵自动选择高频系数。此外,低频子图像的融合是通过平均方法进行的。进一步,融合图像集上的逆优化DT-CWT有助于获得最终的融合医学图像。优化的 DT-CWT 和优化的类型 2 模糊熵的主要目标是最大化 SSIM。实验结果证实,所开发的方法在各种性能测量中优于现有的融合算法。
更新日期:2021-03-18
down
wechat
bug