基于多级残差U-Net的稀疏SPECT图像重建

doi:10.12052/gdutxb.220016

广东工业大学学报 ›› 2023, Vol. 40 ›› Issue (01): 61-67.doi: 10.12052/gdutxb.220016

基于多级残差U-Net的稀疏SPECT图像重建

叶文权, 李斯, 凌捷

广东工业大学计算机学院，广东广州 510006

收稿日期:2022-01-24 出版日期:2023-01-25 发布日期:2023-01-12
通信作者: 李斯(1985-),男,副教授,主要研究方向为医学影像、最优化理论等,E-mail:sili@gdut.edu.cn;凌捷(1964-),男,教授,博士生导师,主要研究方向为网络信息安全技术,E-mail:jling@gdut.edu.cn
作者简介:叶文权(1997-),男,硕士研究生,主要研究方向为深度学习、医学影像等,E-mail:2111905122@mail2.gdut.edu.cn
基金资助:
国家自然科学基金资助项目(11771464)；广东省自然科学基金资助项目(2021A1515012290)；中山大学广东省计算科学重点实验室开放基金资助项目(2021007)

Sparse-view SPECT Image Reconstruction Based on Multilevel-residual U-Net

Ye Wen-quan, Li Si, Ling Jie

School of Computer Science and Technology, Guangdong University of Technology, Guangzhou 510006, China

Received:2022-01-24 Online:2023-01-25 Published:2023-01-12

摘要/Abstract

摘要： 低剂量单光子发射型断层扫描(Single-Photon Emission Computed Tomography，SPECT) 能够减少放射性示踪剂对人体的辐射损害，因此其临床应用变得愈发重要。SPECT扫描可通过投影角度稀疏采样实现低剂量成像；若直接对稀疏采样投影数据进行迭代重建，投影角度的缺失将导致重建图像中出现严重的射线伪影。现今主流的临床方法普遍在图像重建优化模型中引入特定的正则项以抑制射线伪影，然而该类方法不具有通用性，并且正则项过度依赖于经验选取。本文提出一种新颖的神经网络结构以学习稀疏采样投影数据与全角度采样投影数据之间的映射关系；通过所提网络结构合成缺失角度的投影数据，来提升重建图像的质量。数值实验表明，相较于传统迭代重建方法，论文重建方法所得图像的结构相似性(Structural Similarity, SSIM)提高了59%，标准均方误差(Normalized Mean-Square-Error, NMSE)降低了67%，峰值信噪比(Peak Signal-to-Noise Ratio, PSNR)提高了2.48 dB。因此，所提方法能较好地改善稀疏采样投影数据成像后的图像质量。

关键词: 神经网络, SPECT重建, 残差学习, U-Net

Abstract: Low-dose single-photon emission computed tomography (SPECT) imaging can reduce the radiation damage to human bodies caused by radioactive tracers, and hence it is becoming more and more important in clinical practice. In a SPECT system, low-dose imaging can be achieved by acquiring projection data of sparse-view. The sparse-view projection data, if directly reconstructed by conventional iterative reconstruction methods, will inevitably lead to severe ray artifacts in the image domain. Existing clinical reconstruction methods usually introduce specific regularization to the optimization model to suppress ray artifacts. However, this type of methods may not adapt to projection data with various dosage, and the form of regularization heavily depends on prior knowledge. A novel neural network architecture is proposed to learn the mapping from the sparse-view projection data to the full-view projection data. The projection data of missing view angle is synthesized by the proposed neural network to improve the quality of reconstructed images. Numerical experiments show that, compared with the traditional iterative reconstruction method, the SSIM of the reconstructed image is increased by 59%, the NMSE is reduced by 67%, and the PSNR is increased by 2.48 dB. Therefore, the proposed method can better improve the image quality of sparse-view projection data.

Key words: neural network, SPECT reconstruction, residual learning, U-Net

中图分类号:

TP391

叶文权, 李斯, 凌捷. 基于多级残差U-Net的稀疏SPECT图像重建[J]. 广东工业大学学报, 2023, 40(01): 61-67.

Ye Wen-quan, Li Si, Ling Jie. Sparse-view SPECT Image Reconstruction Based on Multilevel-residual U-Net[J]. Journal of Guangdong University of Technology, 2023, 40(01): 61-67.

参考文献

[1] O’MALLEY J, ZIESSMAN H, THRALL J. Nuclear medicine: the requisites [M]. 3rd ed. Maryland: Mosby, 2006.
[2] WELLS R G. Dose reduction is good but it is image quality that matters [J]. Nucl Cardiol, 2018, 27: 238-240.
[3] ZHANG J H, LI S, KROL A, et al. Infimal convolution-based regularization for SPECT reconstruction [J]. Medical Physics, 2018, 45: 5397-5410.
[4] KROL A, LI S, SHEN L X, et al. Preconditioned alternating projection algorithms for maximum a posteriori ECT reconstruction [J]. Inverse Problems, 2012, 28: 115005.
[5] JIANG Y, LI S, XU Y S. A higher-order polynomial method for SPECT reconstruction [J]. IEEE Transactions on Medical Imaging, 2019, 38(5): 1271-1283.
[6] 高俊艳, 刘文印, 杨振国. 结合注意力与特征融合的目标跟踪[J]. 广东工业大学学报, 2019, 36(4): 18-23.
GAO J Y, LIU W Y, YANG Z G. Object tracking combined with attention and feature fusion [J]. Journal of Guangdong University of Technology, 2019, 36(4): 18-23.
[7] ZHANG H M, DONG B. A review on deep learning in medical image reconstruction [J]. Journal of the Operations Research Society of China, 2020, 8: 311-340.
[8] RAMON A J, YANG Y Y, PRETORIUS P H, et al. Improving diagnostic accuracy in low-dose SPECT myocardial perfusion imaging with convolutional denoising networks [J]. IEEE Transactions on Medical Imaging, 2020, 39: 2893-2903.
[9] CHEN H, ZHANG Y, KALRA M K, et al. Low-dose CT with a residual encoder-decoder convolutional neural network [J]. IEEE Transactions on Medical Imaging, 2017, 36: 2524-2535.
[10] ZHANG Z C, LIANG X K, DONG X, et al. A sparse-view CT reconstruction method based on combination of densenet and deconvolution [J]. IEEE Transactions on Medical Imaging, 2018, 37: 1407-1477.
[11] BASTY N, GRAU V. Super resolution of cardiac cine MRI sequences using deep learning[C]// Image Analysis for Moving Organ, Breast, and Thoracic Images. Barcelona: Springer, 2018: 23-31.
[12] 夏皓, 蔡念, 王平, 等. 基于多分辨率学习卷积神经网络的磁共振图像超分辨率重建[J]. 广东工业大学学报, 2020, 37(6): 26-31.
XIA H, CAI N, WANG P, et al. Magnetic resonance image super-resolution via multi-resolution learning [J]. Journal of Guangdong University of Technology, 2020, 37(6): 26-31.
[13] YANG Y, SUN J, LI H B, et al. Deep ADMM-Net for compressive sensing MRI[C]//Proceedings of the 30th International Conference on Neural Information Processing Systems. Barcelona: ACM, 2016: 10-18.
[14] YANG Y, SUN J, LI H B, et al. ADMM-CSNet: a deep learning approach for image compressive sensing [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 42: 521-538.
[15] ADLER J, OKTEM O. Learned primal-dual reconstruction [J]. IEEE Transactions on Medical Imaging, 2018, 37: 1322-1332.
[16] HAEGGSTROEM I, SCHMIDTLEN C R, CAMPANELLA G, et al. DeepPET: a deep encoder-decoder network for directly solving the PET reconstruction inverse problem [J]. Medical Image Analysis, 2018, 54: 253-262.
[17] DONG X, VEKHANDE S, CAO G. Sinogram interpolation for sparse-view micro-CT with deep learning neural network[C]//Physics of Medical Imaging. San Diego: SPIE, 2019: 1094820.
[18] YUAN H Z, JIA J Z, ZHU Z X. SIPID: a deep learning framework for sinogram interpolation and image denoising in low-dose CT reconstruction[C]//2018 IEEE 15th International Symposium on Biomedical Imaging. Washington: IEEE, 2018: 1521-1524.
[19] LEE H, LEE J, KIN H, et al. Deep-neural-network based sinogram synthesis for sparse-view CT image reconstruction [J]. IEEE Transactions on Radiation & Plasma Medical Sciences, 2019, 3: 109-119.
[20] SHIRI I, SHEIKHZADEH P, AY M R. Deep-Fill: deep learning based sinogram domain gap filling in positron emission tomography[EB/OL]. (2019-06-16) [2021-12-20]. https://arxiv.org/abs/1906.07168v1.
[21] RONNEBERGER O, FISCHER P, BROX T. U-Net: convolutional networks for biomedical image segmentation[C]//Medical Image Computing and Computer-Assisted Intervention. Munich: MICCAI, 2015: 234-241.
[22] 黄兴, 杨瑞梅. 结合残差U-Net神经网络和DIP的PET图像降噪[J]. 光学技术, 2021, 47(2): 209-216.
HUANG X, YANG R M. PET image denoising based on residual U-Net neural network and DIP [J]. Optical Technique, 2021, 47(2): 209-216.
[23] 陈钧荣, 林涵阳, 陈羽中. 基于U-Net融合的保留纹理的图像去噪方法[J]. 小型微型计算机系统, 2021, 42(4): 791-797.
CHEN J R, LIN H Y, CHEN Y Z. Texture-preserving image denoising method based on U-Net fusion [J]. Journal of Chinese Computer Systems, 2021, 42(4): 791-797.
[24] MCCANN M T, JIN H K, UNSER M. Deep convolutional neural network for inverse problems in imaging: A Review [J]. IEEE Transactions on Image Processing, 2017, 34: 85-95.
[25] HAN S Y, YOO J, YE C J. Deep residual learning for compressed sensing ct reconstruction via persistent homology analysis[EB/OL]. (2016-11-25) [2021-12-20]. https://arxiv.org/abs/1611.06391.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于多级残差U-Net的稀疏SPECT图像重建

Sparse-view SPECT Image Reconstruction Based on Multilevel-residual U-Net

HTML

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

Metrics

本文评价

推荐阅读 0

[1]	谢国波, 林立, 林志毅, 贺笛轩, 文刚. 基于YOLOv4-MP的绝缘子爆裂缺陷检测方法[J]. 广东工业大学学报, 2023, 40(02): 15-21.
[2]	张锐, 吕俊. 基于分离结果信噪比估计与自适应调频网络的单通道语音分离技术[J]. 广东工业大学学报, 2023, 40(02): 45-54.
[3]	邱俊豪, 程志键, 林国怀, 任鸿儒, 鲁仁全. 具有执行器故障的非线性系统指定性能控制[J]. 广东工业大学学报, 2023, 40(02): 55-63.
[4]	陈靖宇, 吕毅. 基于脉冲神经网络的冷链制冷机结霜检测方法[J]. 广东工业大学学报, 2023, 40(01): 29-38.
[5]	彭美春, 阳晨, 李君平, 叶伟斌, 黄文伟. 基于BP神经网络的车辆碳排放测算研究[J]. 广东工业大学学报, 2023, 40(01): 107-112.
[6]	刘洪伟, 林伟振, 温展明, 陈燕君, 易闽琦. 基于MABM的消费者情感倾向识别模型——以电影评论为例[J]. 广东工业大学学报, 2022, 39(06): 1-9.
[7]	章云, 王晓东. 基于受限样本的深度学习综述与思考[J]. 广东工业大学学报, 2022, 39(05): 1-8.
[8]	彭积广, 肖涵臻. 模型预测控制下多移动机器人的跟踪与避障[J]. 广东工业大学学报, 2022, 39(05): 93-101.
[9]	黎耀东, 任志刚, 吴宗泽. 基于深度神经网络的注塑过程预测控制[J]. 广东工业大学学报, 2022, 39(05): 120-126,136.
[10]	曾江毅, 李志生, 欧耀春, 金宇凯. 季节指数改进的PM_2.5质量浓度组合预测模型研究[J]. 广东工业大学学报, 2022, 39(03): 89-94.
[11]	Gary Yen, 栗波, 谢胜利. 地球流体动力学模型恢复的长短期记忆网络渐进优化方法[J]. 广东工业大学学报, 2021, 38(06): 1-8.
[12]	郭心德, 丁宏强. 离散制造智能工厂场景的AGV路径规划方法[J]. 广东工业大学学报, 2021, 38(06): 70-76.
[13]	黄剑航, 王振友. 基于特征融合的深度学习目标检测算法研究[J]. 广东工业大学学报, 2021, 38(04): 52-58.
[14]	马少鹏, 梁路, 滕少华. 一种轻量级的高光谱遥感图像分类方法[J]. 广东工业大学学报, 2021, 38(03): 29-35.
[15]	夏皓, 蔡念, 王平, 王晗. 基于多分辨率学习卷积神经网络的磁共振图像超分辨率重建[J]. 广东工业大学学报, 2020, 37(06): 26-31.