《计算机应用》唯一官方网站 ›› 2022, Vol. 42 ›› Issue (4): 1308-1316.DOI: 10.11772/j.issn.1001-9081.2021050876
• 多媒体计算与计算机仿真 • 上一篇
李昆鹏1,2, 张鹏程1,2(), 上官宏3, 王燕玲4, 杨婕5, 桂志国1,2
收稿日期:
2021-05-27
修回日期:
2021-09-12
接受日期:
2021-10-14
发布日期:
2021-12-28
出版日期:
2022-04-10
通讯作者:
张鹏程
作者简介:
李昆鹏(1996—),男,河南信阳人,硕士研究生,主要研究方向:医学图像重建、医学图像处理基金资助:
Kunpeng LI1,2, Pengcheng ZHANG1,2(), Hong SHANGGUAN3, Yanling WANG4, Jie YANG5, Zhiguo GUI1,2
Received:
2021-05-27
Revised:
2021-09-12
Accepted:
2021-10-14
Online:
2021-12-28
Published:
2022-04-10
Contact:
Pengcheng ZHANG
About author:
LI Kunpeng, born in 1996, M. S. candidate. His research interests include medical image reconstruction, medical image processing.Supported by:
摘要:
针对采用时域滤波器解析重建后图像存在伪影和图像细节丢失等问题,提出了一种基于卷积神经网络(CNN)的时频域计算机断层扫描(CT)重建算法。首先,在频域中构建了基于卷积神经网络的滤波器网络,实现投影数据的频域滤波;其次,利用反投影操作算子对频域滤波后结果进行域转换得到重建图像;接着,在图像域构建网络对来自反投影层的图像进行处理;最后,在采用最小均方误差损失函数基础上引入多尺度结构相似度损失函数组成复合损失函数,减轻神经网络对结果图像的模糊效应,保留重建图像细节。图像域网络和投影域滤波网络联合作用,最终得到重建结果。在临床数据集上验证了所提算法的有效性,相较于滤波反投影(FBP)算法、全变分(TV)算法及图像域残差编解码CNN(RED-CNN)算法,当投影数目分别为180和90时,所提算法重建结果图像信噪比(PSNR)和结构相似度(SSIM)最高,且归一化均方根误差(NMSE)最小;当投影数目为360时,所提算法仅次于TV算法。实验结果表明,所提算法可以提高CT图像重建图像质量,是一种可行且有效的方法。
中图分类号:
李昆鹏, 张鹏程, 上官宏, 王燕玲, 杨婕, 桂志国. 基于卷积神经网络的时频域CT重建算法[J]. 计算机应用, 2022, 42(4): 1308-1316.
Kunpeng LI, Pengcheng ZHANG, Hong SHANGGUAN, Yanling WANG, Jie YANG, Zhiguo GUI. Time-frequency domain CT reconstruction algorithm based on convolutional neural network[J]. Journal of Computer Applications, 2022, 42(4): 1308-1316.
算法 | 不同投影数目下的耗时 | ||
---|---|---|---|
360 | 180 | 90 | |
FBP | 0.020 6 | 0.014 7 | 0.009 3 |
TV | 1 963.000 0 | 967.000 0 | 495.000 0 |
RED-CNN | 0.064 9 | 0.060 2 | 0.054 5 |
本文算法 | 0.055 4 | 0.038 1 | 0.027 4 |
表1 不同算法单张图像重建耗时 (s)
Tab. 1 Reconstruction of time single image by different algorithms
算法 | 不同投影数目下的耗时 | ||
---|---|---|---|
360 | 180 | 90 | |
FBP | 0.020 6 | 0.014 7 | 0.009 3 |
TV | 1 963.000 0 | 967.000 0 | 495.000 0 |
RED-CNN | 0.064 9 | 0.060 2 | 0.054 5 |
本文算法 | 0.055 4 | 0.038 1 | 0.027 4 |
算法 | 投影数目为360 | 投影数目为180 | 投影数目为90 | ||||||
---|---|---|---|---|---|---|---|---|---|
PSNR/dB | SSIM | NMSE | PSNR/dB | SSIM | NMSE | PSNR/dB | SSIM | NMSE | |
FBP | 15.167 9±0.801 9 | 0.277 0±0.016 0 | 0.721 9±0.218 7 | 12.126 7±0.601 2 | 0.200 2±0.014 4 | 1.441 9±0.397 0 | 9.468 6±0.539 0 | 0.147 2±0.015 3 | 2.643 6±0.703 1 |
TV | 35.954 3±1.207 5 | 0.889 5±0.013 6 | 0.005 9±0.001 4 | 32.259 1±1.408 6 | 0.835 2±0.018 5 | 0.013 9±0.004 2 | 26.268 6±1.352 7 | 0.732 4±0.021 5 | 0.055 8±0.017 8 |
RED-CNN | 34.176 9±1.146 1 | 0.827 9±0.022 8 | 0.009 0±0.003 0 | 31.347 0±0.628 7 | 0.738 3±0.013 1 | 0.017 3±0.004 8 | 28.944 5±1.800 8 | 0.661 9±0.045 5 | 0.034 4±0.034 7 |
本文算法 | 35.505 9±1.129 3 | 0.865 4±0.020 7 | 0.006 5±0.001 6 | 34.267 1±0.779 4 | 0.845 4±0.025 2 | 0.008 7±0.002 3 | 32.119 3±0.677 4 | 0.770 8±0.022 8 | 0.014 2±0.003 3 |
表2 不同算法测试集上指标对比
Tab. 2 Index comparison of different comparison algorithms under different projections
算法 | 投影数目为360 | 投影数目为180 | 投影数目为90 | ||||||
---|---|---|---|---|---|---|---|---|---|
PSNR/dB | SSIM | NMSE | PSNR/dB | SSIM | NMSE | PSNR/dB | SSIM | NMSE | |
FBP | 15.167 9±0.801 9 | 0.277 0±0.016 0 | 0.721 9±0.218 7 | 12.126 7±0.601 2 | 0.200 2±0.014 4 | 1.441 9±0.397 0 | 9.468 6±0.539 0 | 0.147 2±0.015 3 | 2.643 6±0.703 1 |
TV | 35.954 3±1.207 5 | 0.889 5±0.013 6 | 0.005 9±0.001 4 | 32.259 1±1.408 6 | 0.835 2±0.018 5 | 0.013 9±0.004 2 | 26.268 6±1.352 7 | 0.732 4±0.021 5 | 0.055 8±0.017 8 |
RED-CNN | 34.176 9±1.146 1 | 0.827 9±0.022 8 | 0.009 0±0.003 0 | 31.347 0±0.628 7 | 0.738 3±0.013 1 | 0.017 3±0.004 8 | 28.944 5±1.800 8 | 0.661 9±0.045 5 | 0.034 4±0.034 7 |
本文算法 | 35.505 9±1.129 3 | 0.865 4±0.020 7 | 0.006 5±0.001 6 | 34.267 1±0.779 4 | 0.845 4±0.025 2 | 0.008 7±0.002 3 | 32.119 3±0.677 4 | 0.770 8±0.022 8 | 0.014 2±0.003 3 |
算法 | PSNR/dB | SSIM | NMSE |
---|---|---|---|
FBP | 15.158 3±0.761 8 | 0.270 7±0.014 2 | 0.720 4±0.209 5 |
TV | 34.503 8±1.314 3 | 0.842 7±0.019 2 | 0.008 2±0.002 0 |
RED-CNN | 34.087 5±1.100 3 | 0.823 5±0.022 7 | 0.009 2±0.002 9 |
本文算法 | 34.959 6±0.941 8 | 0.843 8±0.020 1 | 0.007 5±0.002 1 |
表3 不同算法在360个投影数目下噪声数据集指标对比
Tab. 3 Index comparison by different algorithms under 360 projections on noisy datasets
算法 | PSNR/dB | SSIM | NMSE |
---|---|---|---|
FBP | 15.158 3±0.761 8 | 0.270 7±0.014 2 | 0.720 4±0.209 5 |
TV | 34.503 8±1.314 3 | 0.842 7±0.019 2 | 0.008 2±0.002 0 |
RED-CNN | 34.087 5±1.100 3 | 0.823 5±0.022 7 | 0.009 2±0.002 9 |
本文算法 | 34.959 6±0.941 8 | 0.843 8±0.020 1 | 0.007 5±0.002 1 |
1 | BRENNER D J, HALL E J. Computed tomography — an increasing source of radiation exposure [J]. New England Journal of Medicine, 2007, 357(22): 2277-2284. 10.1056/nejmra072149 |
2 | SHEPP L A, LOGAN B F. The Fourier reconstruction of a head section [J]. IEEE Transactions on Nuclear Science, 1974, 21(3): 21-43. 10.1109/tns.1974.6499235 |
3 | BATENBURG K J, PLANTAGIE L. Fast approximation of algebraic reconstruction methods for tomography [J]. IEEE Transactions on Image Processing, 2012, 21(8): 3648-3658. 10.1109/tip.2012.2197012 |
4 | DING Q Q, LONG Y, ZHANG X Q, et al. Modeling mixed poisson-gaussian noise in statistical image reconstruction for X-ray CT [C]// Proceedings of the 4th International Conference on Image Formation in X-Ray Computed Tomography. Piscataway: IEEE, 2018: 399-402. |
5 | YE S Q, RAVISHANKAR S, LONG Y, et al. Adaptive sparse modeling and shifted-poisson likelihood based approach for low-dose CT image reconstruction [C]// Proceedings of the 2017 IEEE 27th International Workshop on Machine Learning for Signal Processing. Piscataway: IEEE, 2017: 1-6. 10.1109/mlsp.2017.8168124 |
6 | DING Q Q, LONG Y, ZHANG X Q, et al. Statistical image reconstruction using mixed Poisson-Gaussian noise model for X-ray CT [EB/OL]. [2018-01-19]. . |
7 | YE S Q, RAVISHANKAR S, LONG Y, et al. SPULTRA: Low-dose CT image reconstruction with joint statistical and learned image models [J]. IEEE Transactions on Medical Imaging, 2020, 39(3): 729-741. 10.1109/tmi.2019.2934933 |
8 | SIDKY E Y, PAN X C. Image reconstruction in circular cone-beam computed tomography by constrained, total-variation minimization[J]. Physics in Medicine & Biology, 2008, 53(17): 4777-4807. 10.1088/0031-9155/53/17/021 |
9 | NIU S Z, GAO Y, BIAN Z Y, et al. Sparse-view X-ray CT reconstruction via total generalized variation regularization [J]. Physics in Medicine & Biology, 2014, 59(12): 2997. 10.1088/0031-9155/59/12/2997 |
10 | ZHANG H, HUANG J, MA J H, et al. Iterative reconstruction for X-ray computed tomography using prior-image induced nonlocal regularization [J]. IEEE Transactions on Biomedical Engineering, 2014, 61(9): 2367-2378. 10.1109/tbme.2013.2287244 |
11 | XU Q, YU H, MOU X, et al. Low-dose X-ray CT reconstruction via dictionary learning[J]. IEEE Transactions on Medical Imaging, 2012, 31(9): 1682-1697. 10.1109/tmi.2012.2195669 |
12 | PALENSTIJN W J, BATENBURG K J, SIJBERS J. Performance improvements for iterative electron tomography reconstruction using Graphics Processing Units (GPUs) [J]. Journal of Structural Biology, 2011, 176(2): 250-253. 10.1016/j.jsb.2011.07.017 |
13 | ZHOU X Y, WEI G, FU W L, et al. Application of deep learning in object detection [C]// Proceedings of the 2017 IEEE/ACIS 16th International Conference on Computer and Information Science. Piscataway: IEEE, 2017: 631-634. 10.1109/icis.2017.7960069 |
14 | LIU W B, WANG Z D, LIU X H, et al. A survey of deep neural network architectures and their applications[J]. Neurocomputing, 2017, 234: 11-26. 10.1016/j.neucom.2016.12.038 |
15 | LECUN Y, BENGIO Y, HINTON G. Deep learning [J]. Nature, 2015, 521(7553): 436-444. 10.1038/nature14539 |
16 | ZHU B, LIU J Z, CAULEY S F, et al. Image reconstruction by domain-transform manifold learning [J]. Nature, 2018, 555(7697): 487-492. 10.1038/nature25988 |
17 | LI Y S, LI K, ZHANG C Z, et al. Learning to reconstruct Computed Tomography (CT) images directly from sinogram data under a variety of data acquisition conditions[J]. IEEE Transactions on Medical Imaging, 2019, 38(10): 2469-2481. 10.1109/tmi.2019.2910760 |
18 | WU D F, KIM K S, FAKHRI G E, et al. Iterative low-dose CT reconstruction with priors trained by artificial neural network [J]. IEEE Transactions on Medical Imaging, 2017, 36(12): 2479-2486. 10.1109/tmi.2017.2753138 |
19 | CHEN H, ZHANG Y, CHEN Y J, et al. LEARN: learned experts’ assessment-based reconstruction network for sparse-data CT [J]. IEEE Transactions on Medical Imaging, 2018, 37(6): 1333-1347. 10.1109/tmi.2018.2805692 |
20 | GUPTA H, JIN K H, NGUYEN H Q, et al. CNN-based projected gradient descent for consistent image reconstruction[J]. IEEE Transactions on Medical Imaging, 2018, 37(6): 1440-1453. 10.1109/tmi.2018.2832656 |
21 | WANG J, ZENG L, WANG C, et al. ADMM-based deep reconstruction for limited-angle CT [J]. Physics in Medicine & Biology, 2019, 64(11): 115011. 10.1088/1361-6560/ab1aba |
22 | HE J, YANG Y, WANG Y B, et al. Optimizing a parameterized plug-and-play ADMM for iterative low-dose CT reconstruction [J]. IEEE Transactions on Medical Imaging, 2019, 38(2): 371-382. 10.1109/tmi.2018.2865202 |
23 | DING Q, CHEN G, ZHANG X, et al. Low-dose CT with deep learning regularization via proximal forward-backward splitting[J]. Physics in Medicine & Biology, 2020, 65(12): 125009. 10.1088/1361-6560/ab831a |
24 | ZHANG Q, HU Z, JIANG C, et al. Artifact removal using a hybrid-domain convolutional neural network for limited-angle computed tomography imaging[J]. Physics in Medicine & Biology, 2020, 65(15): 155010. 10.1088/1361-6560/ab9066 |
25 | ZHENG A, GAO H, ZHANG L, et al. A dual-domain deep learning-based reconstruction method for fully 3D sparse data helical CT [J]. Physics in Medicine & Biology, 2020, 65(24): 245030. 10.1088/1361-6560/ab8fc1 |
26 | GE Y, SU T, ZHU J, et al. ADAPTIVE-NET: deep computed tomography reconstruction network with analytical domain transformation knowledge [J]. Quantitative Imaging in Medicine and Surgery, 2020, 10(2): 415-427. 10.21037/qims.2019.12.12 |
27 | LIANG K C, YANG H K, XING Y X. Comparison of projection domain, domainimage, and comprehensive deep learning for sparse-view X-ray CT image reconstruction [EB/OL]. [2019-5-28]. . |
28 | LIN W A, LIAO H F, PENG C, et al. DuDoNet: dual domain network for CT metal artifact reduction [C]// Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2019: 10504-10513. 10.1109/cvpr.2019.01076 |
29 | WÜRFL T, HOFFMANN M, CHRISTLEIN V, et al. Deep learning computed tomography: learning projection-domain weights from image domain in limited angle problems[J]. IEEE Transactions on Medical Imaging, 2018, 37(6): 1454-1463. 10.1109/tmi.2018.2833499 |
30 | HE J, WANG Y, MA J. Radon inversion via deep learning [J]. IEEE Transactions on Medical Imaging, 2020, 39(6): 2076-2087. 10.1109/tmi.2020.2964266 |
31 | ZENG G L. A filtered backprojection MAP algorithm with nonuniform sampling and noise modeling [J]. Medical Physics, 2012, 39(4): 2170-2178. 10.1118/1.3697736 |
32 | PELT D M, BATENBURG K J. Improving filtered backprojection reconstruction by data-dependent filtering [J]. IEEE Transactions on Image Processing, 2014, 23(11): 4750-4762. 10.1109/tip.2014.2341971 |
33 | 曾更生. 医学图像重建[M]. 北京:高等教育出版社, 2009: 22-24. 10.1190/1.3603805 |
ZENG G S. Medical Image Reconstruction [M]. Beijing: Higher Education Press, 2009: 22-24. 10.1190/1.3603805 | |
34 | 庄天戈. CT原理与算法[M]. 上海: 上海交通大学出版社, 1992:38-48. 10.1007/978-1-4471-3275-2_19 |
ZHUANG T G. CT Principle and Algorithm [M]. Shanghai: Shanghai Jiao Tong University Press, 1992: 38-48. 10.1007/978-1-4471-3275-2_19 | |
35 | HU C, YI Z, KALRA M K, et al. Low-dose CT with a Residual Encoder-Decoder Convolutional Neural Network (RED-CNN) [J]. IEEE Transactions on Medical Imaging, 2017, 36(12): 2524-2535. 10.1109/tmi.2017.2715284 |
36 | GHANI M U, KARL W C. CNN based sinogram denoising for low-dose CT [C]// Proceedings of the 2018 International Conference on Imaging and Applied Optics. Oriando,FL: OPTICA Publishing Group, 2018: MM2D.5. 10.1364/math.2018.mm2d.5 |
37 | WHITELEY W, GREGOR J. CNN-based PET sinogram repair to mitigate defective block detectors [J]. Physics in Medicine and Biology, 2019, 64(23): 235017. 10.1088/1361-6560/ab4919 |
38 | WANG B, LIU H. FBP-Net for direct reconstruction of dynamic PET images [J]. Physics in Medicine & Biology, 2020, 65(23): 235008. 10.1088/1361-6560/abc09d |
39 | RONNEBERGER O, FISCHER P, BROX T. U-Net: convolutional networks for biomedical image segmentation [C]// Proceedings of the 2015 International Conference on Medical Image Computing and Computer-Assisted Intervention. Cham: Springer, 2015: 234-241. 10.1007/978-3-319-24574-4_28 |
40 | ZHAO H, GALLO O, FROSIO I, et al. Loss functions for image restoration with neural networks [J]. IEEE Transactions on Computational Imaging, 2017,3(1): 47-57. 10.1109/tci.2016.2644865 |
41 | XIA W, LU Z, HUANG Y, et al. MAGIC: manifold and graph integrative convolutional network for low-dose CT reconstruction [J].IEEE Transactions on Medical Imaging, 2021,40(12): 3459-3472. 10.1109/tmi.2021.3088344 |
42 | BIGURI A, DOSANJH M, HANCOCK S, et al. TIGRE: a MATLAB-GPU toolbox for CBCT image reconstruction [J]. Biomedical Physics & Engineering Express, 2016, 2(5): 055010. 10.1088/2057-1976/2/5/055010 |
[1] | 秦璟, 秦志光, 李发礼, 彭悦恒. 基于概率稀疏自注意力神经网络的重性抑郁疾患诊断[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2970-2974. |
[2] | 李云, 王富铕, 井佩光, 王粟, 肖澳. 基于不确定度感知的帧关联短视频事件检测方法[J]. 《计算机应用》唯一官方网站, 2024, 44(9): 2903-2910. |
[3] | 陈虹, 齐兵, 金海波, 武聪, 张立昂. 融合1D-CNN与BiGRU的类不平衡流量异常检测[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2493-2499. |
[4] | 赵宇博, 张丽萍, 闫盛, 侯敏, 高茂. 基于改进分段卷积神经网络和知识蒸馏的学科知识实体间关系抽取[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2421-2429. |
[5] | 张春雪, 仇丽青, 孙承爱, 荆彩霞. 基于两阶段动态兴趣识别的购买行为预测模型[J]. 《计算机应用》唯一官方网站, 2024, 44(8): 2365-2371. |
[6] | 高阳峄, 雷涛, 杜晓刚, 李岁永, 王营博, 闵重丹. 基于像素距离图和四维动态卷积网络的密集人群计数与定位方法[J]. 《计算机应用》唯一官方网站, 2024, 44(7): 2233-2242. |
[7] | 王东炜, 刘柏辰, 韩志, 王艳美, 唐延东. 基于低秩分解和向量量化的深度网络压缩方法[J]. 《计算机应用》唯一官方网站, 2024, 44(7): 1987-1994. |
[8] | 沈君凤, 周星辰, 汤灿. 基于改进的提示学习方法的双通道情感分析模型[J]. 《计算机应用》唯一官方网站, 2024, 44(6): 1796-1806. |
[9] | 姚迅, 秦忠正, 杨捷. 生成式标签对抗的文本分类模型[J]. 《计算机应用》唯一官方网站, 2024, 44(6): 1781-1785. |
[10] | 黄梦源, 常侃, 凌铭阳, 韦新杰, 覃团发. 基于层间引导的低光照图像渐进增强算法[J]. 《计算机应用》唯一官方网站, 2024, 44(6): 1911-1919. |
[11] | 李健京, 李贯峰, 秦飞舟, 李卫军. 基于不确定知识图谱嵌入的多关系近似推理模型[J]. 《计算机应用》唯一官方网站, 2024, 44(6): 1751-1759. |
[12] | 席治远, 唐超, 童安炀, 王文剑. 基于双路时空网络的驾驶员行为识别[J]. 《计算机应用》唯一官方网站, 2024, 44(5): 1511-1519. |
[13] | 高文烁, 陈晓云. 基于节点结构的点云分类网络[J]. 《计算机应用》唯一官方网站, 2024, 44(5): 1471-1478. |
[14] | 魏凤凤, 陈伟能. 分布式数据驱动的多约束进化优化算法[J]. 《计算机应用》唯一官方网站, 2024, 44(5): 1393-1400. |
[15] | 孙敏, 成倩, 丁希宁. 基于CBAM-CGRU-SVM的Android恶意软件检测方法[J]. 《计算机应用》唯一官方网站, 2024, 44(5): 1539-1545. |
阅读次数 | ||||||
全文 |
|
|||||
摘要 |
|
|||||