北京生物醫(yī)學(xué)工程

基于醫(yī)學(xué)圖像的超分辨率重建算法綜述

Review of super-resolution reconstruction algorithms in medical image

作者：成云鳳汪偉

單位：上海理工大學(xué)（上海 200093）

關(guān)鍵詞：超分辨率重建；醫(yī)學(xué)圖像處理；字典學(xué)習(xí)；稀疏表示；卷積神經(jīng)網(wǎng)絡(luò)

分類號(hào)：R318.04

出版年·卷·期（頁(yè)碼）：2019·38·5（535-543）

摘要：

隨著臨床對(duì)醫(yī)學(xué)圖像高分辨率的要求，基于低分辨率醫(yī)學(xué)圖像的超分辨率重建算法已成為研究熱點(diǎn)，該類方法在不需要改進(jìn)硬件設(shè)備的情況下，可以顯著提高圖像分辨率，因此對(duì)其進(jìn)行綜述具有重要意義。針對(duì)醫(yī)學(xué)圖像領(lǐng)域中特有的超分辨率重建算法，首先分析了該類算法的研究現(xiàn)狀，并將其分為三類:基于插值的超分辨率重建、基于重構(gòu)的超分辨率重建和基于學(xué)習(xí)的超分辨率重建。同時(shí)，基于MR圖像、CT圖像、超聲圖像等細(xì)分醫(yī)學(xué)圖像領(lǐng)域，深入分析了超分辨率重建算法的研究進(jìn)展，并對(duì)不同類型的算法進(jìn)行了歸納總結(jié)和比對(duì)分析。其次，對(duì)超分辨率重建算法所對(duì)應(yīng)的評(píng)價(jià)標(biāo)準(zhǔn)也進(jìn)行了介紹。最后，展望了超分辨率重建技術(shù)在醫(yī)學(xué)圖像領(lǐng)域的發(fā)展趨勢(shì)。當(dāng)前應(yīng)用于醫(yī)學(xué)圖像領(lǐng)域的超分辨重建算法已經(jīng)發(fā)展到一定水平，已經(jīng)逐步突破基于單一方法的研究形式，通過(guò)與機(jī)器學(xué)習(xí)與稀疏表示等理論的深度融合，形成了更高效的算法。

With the demand of high resolution in clinical medical image, the super-resolution reconstruction algorithm based on low-resolution medical images has become a research hotspot. This method can significantly improve the image resolution without the need of upgrading hardware devices, so it is of great significance to review them. In terms of the special super-resolution reconstruction algorithms in the field of medical images, this paper firstly analyzes the research status of related algorithms, then divides them into three categories: the algorithm based on interpolation, reconfiguration algorithms and learning algorithms. Meanwhile, based on the subdivision of medical image field, including MR image, CT image and ultrasonic image, the research progress of super-resolution reconstruction algorithm is deeply analyzed, and different types of algorithms are summarized and compared. Secondly, the evaluation criteria of the super-resolution reconstruction algorithm are also introduced. Finally, the development trend of super-resolution reconstruction technology in the field of medical imaging is prospected. At present, these kinds of technologies have been developed rapidly, and infused with other methods, such as machine learning and sparse representation theories, which may bring out more efficient algorithms.

參考文獻(xiàn)：

[1] 徐軍,劉慧,尹義龍.基于非局部自回歸學(xué)習(xí)的醫(yī)學(xué)圖像超分辨重建方法[J].模式識(shí)別與人工智能,2017,30(8):747-753.

Xu J, Liu H, Yin YL. A method of medical image super-resolution reconstruction based on non-local autoregressive learning[J]. Pattern Recognition and Artificial Intelligence , 2017,30(8):747-753.

[2] 庾吉飛.基于學(xué)習(xí)的圖像超分辨率重建算法研究[D].西安：西安電子科技大學(xué)，2013.

Yu JF. Research on image super-resolution reconstruction algorithm based on learning[D]. Xi’an ：Xidian University，2013.

[3] 袁晶.基于超分辨重建數(shù)學(xué)模型的醫(yī)學(xué)圖像處理方法[J].計(jì)算機(jī)仿真,2014,31(5):243-245.

Yuan J. Medical image processing method based on super-resolution reconstruction mathematical model[J]. Computer Simulation ,2014,31(5):243-245.

[4] Tsai RY，Huang TS. Multi-frame image restoration and registration[J]. Advance Computer Visual and Image Processing, 1984, 1: 317-339.

[5] Mahmoudzadeh AP, Kashou NH. Interpolation-based super-resolution reconstruction: effects of slice thickness[J]. Journal of Medical Imaging, 2014, 1(3): 034007.

[6] Luo J, Mou Z, Qin B, et al. Fast single image super-resolution using estimated low-frequency k-space data in MRI[J]. Magnetic resonance imaging, 2017, 40: 1-11.

[7] van Aarle W, Batenburg KJ, Van Gompel G, et al. Super-resolution for computed tomography based on discrete tomography[J]. IEEE Transactions on image processing, 2014, 23(3): 1181-1193.

[8] Mahmoud A, Taher F, Al-Ahmad H. Two dimensional filters for enhancing the resolution of interpolated CT scan images[C]//Innovations in Information Technology (IIT), 2016 12th International Conference on IEEE, 2016: 1-6.

[9] 張煜,吳秀秀,戴振暉. 基于配準(zhǔn)的肺4D-CT圖像超分辨率重建方法:中國(guó), 103886568[P].2014-06-25.

[10] Morin R, Basarab A, Bidon S, et al. Motion estimation-based image enhancement in ultrasound imaging[J]. Ultrasonics, 2015, 60: 19-26.

[11] Huang Q, Huang Y, Hu W, et al. Bezier interpolation for 3-D freehand ultrasound[J]. IEEE Transactions on Human-Machine Systems, 2015, 45(3): 385-392.

[12] Chang G, Pan T, Qiao F, et al. Comparison between two super-resolution implementations in PET imaging[J]. Medical physics, 2009, 36(4): 1370-1383.

[13] Farsiu S, Robinson MD, Elad M, et al. Fast and robust multiframe super resolution[J]. IEEE transactions on image processing, 2004, 13(10): 1327-1344.

[14] Irani M, Peleg S. Super resolution from image sequences[C]//[1990] Proceedings. 10th International Conference on Pattern Recognition. IEEE, 1990, 2: 115-120.

[15] Stark H, Olsen E T. Projection-based image restoration[J]. JOSA A, 1992, 9(11): 1914-1919.

[16] Schultz RR, Stevenson RL. Motion-compensated scan conversion of interlaced video sequences[C]//Image and Video Processing IV. International Society for Optics and Photonics, 1996, 2666: 107-119.

[17] Zheng H, Zeng K, Guo D, et al. Multi-Contrast Brain MRI Image Super-Resolution With Gradient-Guided Edge Enhancement[J]. IEEE Access, 2018, 6: 57856-57867.

[18] Viermetz MP, Birnbacher LJB, Fehringer A, et al. High resolution laboratory grating-based X-ray phase-contrast CT[C]//Medical Imaging 2017: Physics of Medical Imaging. International Society for Optics and Photonics, 2017, 10132: 101325K.

[19] Li Y, Matej S, Metzler SD. Image reconstructions from super-sampled data sets with resolution modeling in PET imaging[J]. Medical physics, 2014, 41(12)：12912-1-12912-14.

[20] Malczewski K, Buczkowski M. Semi-propeller compressed sensing MR image super-resolution reconstruction[C]//Signals and Electronic Systems (ICSES), 2014 International Conference on. IEEE, 2014: 1-4.

[21] Xu H, Miao H, Yang C, et al. Research on super-resolution image reconstruction based on an improved POCS algorithm[C]//International Conference on Optical and Photonic Engineering (icOPEN 2015). International Society for Optics and Photonics, 2015, 9524: 95242V.

[22] 李宇宙, 張喆, 陳泉榮, 等. 一種以超分辨率理論為基礎(chǔ)的磁共振眼球成像方法[J]. 波譜學(xué)雜志, 2017, 34(4): 439-452.

Li YZ, Zhang Z, Chen QR, et al. A magnetic resonance eye imaging method based on super - resolution theory[J]. Chinese Journal of Magnetic Resonance , 2017, 34(4): 439-452.

[23] Han XH, Iwamoto Y, Shiino A, et al. Robust isotropic super-resolution by maximizing a Laplace posterior for MRI volumes[C]//Medical Imaging 2014: Image Processing. International Society for Optics and Photonics, 2014, 9034: 90342I.

[24] Zhao N, Wei Q, Basarab A, et al. Single image super-resolution of medical ultrasound images using a fast algorithm[C]//2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI). IEEE, 2016: 473-476.

[25] Freeman WT, Jones TR, Pasztor EC. Example-based super-resolution[J]. IEEE Computer graphics and Applications, 2002, 22(2): 56-65.

[26] Yang J, Wright J, Huang T, et al. Image super-resolution as sparse representation of raw image patches[C]// IEEE Conference on on Computer Vision and Pattern Recognition .Anchorage, AK：CVPR，2008： 1–8.

[27] 李欣, 崔子冠, 朱秀昌. 超分辨率重建算法綜述[J]. 電視技術(shù), 2016, 40(9): 1-9.

Li X,Cui ZG,Zhu XC. Overview of super resolution reconstruction algorithms[J]. Television Technology, 2016, 40(9): 1-9.

[28] Bustin A, Voilliot D, Menini A, et al. Isotropic reconstruction of mr images using 3D patch-based self-similarity learning[J]. IEEE Transactions on Medical Imaging, 2018, 37(8): 1932-1942.

[29] Ota J, Umehara K, Ishimaru N, et al. Evaluation of the sparse coding super-resolution method for improving image quality of up-sampled images in computed tomography[C]//Medical Imaging 2017: Image Processing. International Society for Optics and Photonics, 2017, 10133: 101331S.

[30] Zeng XH, Hou SL. Manifold-regularization super-resolution image reconstruction[J]. Journal of Computers, 2017, 28(1): 119-136.

[31] Yang G, Ye X, Slabaugh G, et al. Super-resolved enhancement of a single image and its application in cardiac MRI[C]//International Conference on Image and Signal Processing. Berlin：Springer, 2016: 179-190.

[32] 席志紅,曾繼琴,李爽.基于雙字典和稀疏表示的醫(yī)學(xué)圖像超分辨率重建[J].計(jì)算機(jī)測(cè)量與控制,2017,25(3):197-200.

Xi ZH,Zeng JQ,Li S. Medical image super-resolution reconstruction based on double dictionary and sparse representation[J]. Computer Measurement & Control , 2017,25(3):197-200.

[33] 梁海蘭. 乳腺 MR 圖像三維超分辨率重建[D]. 天津：天津大學(xué), 2017.

Liang HL. Super-resolution reconstruction of three-dimensional breast MR images[D]. Tianjin ：Tianjin University ,2017.

[34] Asif M, Akram MU, Hassan T, et al. High resolution OCT image generation using super resolution via sparse representation[C]//Eighth International Conference on Graphic and Image Processing (ICGIP 2016). International Society for Optics and Photonics, 2017, 10225: 1022512.

[35] Jiang C, Zhang Q, Fan R, et al. Super-resolution CT Image Reconstruction Based on Dictionary Learning and Sparse Representation[J]. Scientific reports, 2018, 8(1): 8799.

[36] 朱雪茹,李勇明,李傳明,等.基于雙層字典學(xué)習(xí)的低劑量CT圖像重建算法[J].北京生物醫(yī)學(xué)工程,2017, 36(6):584-590.

Zhu XR, Li YM, Li CM, et al. Low dose CT image reconstruction algorithm based on bilayer dictionary learning[J]. Beijing Biomedical Engineering,2017, 36 (6):584-590.

[37] Lai R, Li J. Manifold based low-rank regularization for image restoration and semi-supervised learning[J]. Journal of Scientific Computing, 2018, 74(3): 1241-1263.

[38] 李斌, 李德來(lái), 張瓊. 基于稀疏表示的 B 型超聲圖像超分辨重建算法[J]. 中國(guó)醫(yī)療器械信息, 2014, 20(7): 4-7.

Li B,Li DL,Zhang Q. A super - resolution reconstruction algorithm based on sparse representation of B - type ultrasonic images[J]. China Medical Device Information , 2014, 20(7): 4-7.

[39] Dong C, Loy C C, He K, et al. Learning a deep convolutional network for image super-resolution[C] // European conference on computer vision. Springer, Cham, 2014: 184-199.

[40] 邢曉羊，魏敏，符穎.基于特征損失的醫(yī)學(xué)圖像超分辨率重建[J].計(jì)算機(jī)工程與應(yīng)用，2018,54（20）：202-207.

Xing XY,Wei M,Fu Y. Super-resolution reconstruction of medical images based on feature loss[J]. Computer Engineering and Applications , 2018,54（20）：202-207.

[41] 高媛,劉志,秦品樂(lè),等.基于深度殘差生成對(duì)抗網(wǎng)絡(luò)的醫(yī)學(xué)影像超分辨率算法[J].計(jì)算機(jī)應(yīng)用,2018,9:2689-2695.

Gao Y,Liu Z,Qin PL, et al. Super-resolution algorithm of medical image based on deep residuals to generate adversarial network[J]. Journal of Computer Applications , 2018, 9:2689-2695.

[42] Oktay O, Ferrante E, Kamnitsas K, et al. Anatomically constrained neural networks (ACNNs): application to cardiac image enhancement and segmentation[J]. IEEE transactions on medical imaging, 2018, 37(2): 384-395.

[43] Li H, Gao Y, Dong J, et al. Super-Resolution Based on Noise Resistance Deep Convolutional Network[C]// Proceedings of the 2018 6th International Conference on Bioinformatics and Computational Biology. Switzerland ：ACM, 2018: 88-94.

[44] Caballo M, Fedon C, Brombal L, et al. Development of 3D patient-based super-resolution digital breast phantoms using machine learning[J]. Physics in Medicine & Biology, 2018, 63(22): 225017.

[45] Hong X, Zan Y, Weng F, et al. Enhancing the Image Quality via Transferred Deep Residual Learning of Coarse PET Sinograms[J]. IEEE transactions on medical imaging, 2018, 37(10): 2322-2332.

服務(wù)與反饋：

【文章下載】【加入收藏】

提示：您還未登錄，請(qǐng)登錄！點(diǎn)此登錄

51黑料吃瓜在线观看,51黑料官网|51黑料捷克街头搭讪_51黑料入口最新视频