摘要
视觉Transformer在提升图像恢复性能方面表现出了良好潜能.研究了一种基于小波分解U型Transformer网络的加速磁共振图像(MRI)重建方法.重建网络的核心单元结合了Swin Transformer与Unet结构,通过融合图像的多尺度特征,改进网络的学习能力,以达到更好的重建性能.采用小波变换对输入图像进行分解,减少了Swin Transformer输入的特征维度,从而有效降低了重构网络的计算复杂度.通过引入小波域损失来约束网络的训练,更好地恢复图像的结构纹理信息.在Calgary-Campinas大脑MR数据集上进行实验比较,结果验证了此方法在提升重构图像质量及控制系统计算复杂度方面的有效性.
Transformer has shown great potential to improve image recovery performance.A reconstruction method of the accelerated magnetic resonance image(MRI)based on a wavelet-decomposed U-shaped Transformer network is studied.The core unit of the reconstruction network is designed based on combining the Swin Transformer with the Unet.By fusing multi-scale features of images,the learning ability of the network is improved and better reconstruction performance is obtained.By using wavelet transform to decompose the input image,the input feature dimension of Swin Transformer is reduced,so as to effectively reduce the computational complexity of the reconstruction network.The wavelet domain loss is adopted to constrain network training,with better recovering the structure and texture information of the image.The experimental results on Calgary-Campinas brain MR Dataset verify the effectiveness of the proposed method in improving the quality of reconstructed images and balancing the complexity of system.
作者
熊承义
李帆
高志荣
孙清清
陈文旗
XIONG Chengyi;LI Fan;GAO Zhirong;SUN Qingqing;CHEN Wenqi(South-Center Minzu University,College of Electronic and Information Engineering,Wuhan 430074,China;South-Center Minzu University,Hubei Key Lab of Intelligent Wireless Communication,Wuhan 430074,China;South-Center Minzu University,College of Computer Science,Wuhan 430074,China)
出处
《中南民族大学学报(自然科学版)》
2025年第5期695-702,共8页
Journal of South-Central Minzu University(Natural Science Edition)
基金
多谱信息处理技术国家重点实验室基金资助项目(6142113210303)
中央高校基本科研业务费专项资金资助项目(CZY21013)。
作者简介
熊承义(1969-),男,教授,研究方向:图像处理与模式识别,E-mail:xiongcy@mail.scuec.edu.cn;通信作者:李帆,研究方向:磁共振图像重构,E-mail:731369073@qq.com。
