自动化学报2023，Vol.49Issue(12)：2605-2614,10.DOI:10.16383/j.aas.c201036
融合MRI信息的PET图像去噪:基于图小波的方法
PET Images Denoising With MRI Information:A Graph Wavelet Based Method
摘要
Abstract
Positron emission tomography(PET)is a powerful nuclear medicine functional imaging modality,which is widely used in clinical diagnosis.However,the PET images have low spatial resolution and contain noise.It is ne-cessary to improve the quality of PET images by denoising.As hybrid imaging devices such as PET/MR(magnetic resonance)have become available,magnetic resonance imaging(MRI)prior information could be used to improve the denoising quality of PET images.For dynamic PET images,this paper presents a novel denoising method based on graph wavelet transform(GWT)with MRI information.Firstly,the composite image is generated by the dynam-ic PET frames.Then,we fuse the MRI image and the composite image through the hard threshold method to ob-tain a new fusion image.Next,the graph Laplacian matrix is constructed based on the fusion image.Finally,we perform GWT on the dynamic PET images to denoise.The simulation experiments show that,compared with graph filtering and GWT,and other MRI incorporated methods,the proposed approach has higher SNR(signal-to-noise ratio)while preserving the image lesions details.Compared with VGG(Visual Geometry Group)deep neural network based method,the proposed method has the similar denoising performance,but it does not need a lot of data training and has low computational complexity.关键词
正电子发射断层成像/磁共振成像/图小波/去噪Key words
Positron emission tomography(PET)/magnetic resonance imaging(MRI)/graph wavelet/denoising引用本文复制引用
易利群,盛玉霞,柴利..融合MRI信息的PET图像去噪:基于图小波的方法[J].自动化学报,2023,49(12):2605-2614,10.基金项目
国家自然科学基金(62173259,61625305)资助Supported by National Natural Science Foundation of China(62173259,61625305) （62173259,61625305）
