2.793

                    2018影響因子

                    (CJCR)

                    • 中文核心
                    • EI
                    • 中國科技核心
                    • Scopus
                    • CSCD
                    • 英國科學文摘

                    留言板

                    尊敬的讀者、作者、審稿人, 關于本刊的投稿、審稿、編輯和出版的任何問題, 您可以本頁添加留言。我們將盡快給您答復。謝謝您的支持!

                    姓名
                    郵箱
                    手機號碼
                    標題
                    留言內容
                    驗證碼

                    基于自適應分數階的醫學圖像非剛性配準

                    張桂梅 胡強 郭黎娟

                    張桂梅, 胡強, 郭黎娟. 基于自適應分數階的醫學圖像非剛性配準. 自動化學報, 2020, 46(9): 1941?1951 doi: 10.16383/j.aas.c190027
                    引用本文: 張桂梅, 胡強, 郭黎娟. 基于自適應分數階的醫學圖像非剛性配準. 自動化學報, 2020, 46(9): 1941?1951 doi: 10.16383/j.aas.c190027
                    Zhang Gui-Mei, Hu Qiang, Guo Li-Juan. Medical image non-rigid registration based on adaptive fractional order. Acta Automatica Sinica, 2020, 46(9): 1941?1951 doi: 10.16383/j.aas.c190027
                    Citation: Zhang Gui-Mei, Hu Qiang, Guo Li-Juan. Medical image non-rigid registration based on adaptive fractional order. Acta Automatica Sinica, 2020, 46(9): 1941?1951 doi: 10.16383/j.aas.c190027

                    基于自適應分數階的醫學圖像非剛性配準

                    doi: 10.16383/j.aas.c190027
                    基金項目: 國家自然科學基金(61462065, 61661036)資助
                    詳細信息
                      作者簡介:

                      張桂梅:江西省圖像處理與模式識別重點實驗室(南昌航空大學)教授. 主要研究方向為計算機視覺,圖像處理與模式識別. 本文通信作者.E-mail: guimei.zh@163.com

                      胡強:江西省圖像處理與模式識別重點實驗室(南昌航空大學)碩士研究生. 主要研究方向為圖像處理與計算機視覺.E-mail: 18070517681@163.com

                      郭黎娟:江西省圖像處理與模式識別重點實驗室(南昌航空大學)碩士研究生. 主要研究方向為圖像處理與計算機視覺.E-mail: 13576030184@163.com

                    Medical Image Non-rigid Registration Based on Adaptive Fractional Order

                    Funds: Supported by National Natural Science Foundation of China (61462065, 61661036)
                    • 摘要: 現有的醫學圖像配準算法對于灰度均勻、弱邊緣以及弱紋理圖像易陷入局部最優從而導致配準精度低下、收斂速度緩慢. 分數階主動Demons (Fractional active Demons, FAD)算法是解決該問題的有效方法, 并且適用于圖像的非剛性配準. 但FAD中的最佳分數階階次是人工交互選取, 并且對整幅圖像都是固定不變的. 為了解決該問題, 提出一種階次自適應的主動Demons算法并將其應用到醫學圖像的非剛性配準中. 算法首先根據圖像的局部特征建立分數階階次自適應的數學模型, 并逐像素計算最優階次, 基于該階次構造Riemann-Liouvill (R-L)分數階微分動態模板; 然后將自適應R-L分數階微分引入到Active Demons算法, 在一定程度上緩解了圖像配準在弱邊緣和弱紋理區域易陷入局部最優問題, 從而提高了配準精度. 通過在兩個醫學圖像庫上進行實驗驗證, 實驗結果表明該方法可以處理灰度均勻、弱紋理和弱邊緣的醫學圖像非剛性配準, 配準精度得到較大提升.
                    • 圖  1  反正切函數圖像

                      Fig.  1  Arctan function

                      圖  2  不同切片層的BrainWeb圖像配準結果

                      Fig.  2  Registration result of BrainWeb image of different slice layers

                      圖  3  冠狀面配準結果圖

                      Fig.  3  Regristration results of coronal plane

                      圖  4  矢狀面配準結果圖

                      Fig.  4  Regristration results of sagittal plane

                      圖  5  橫切面配準結果圖

                      Fig.  5  Regristration results of transverse plane

                      表  1  均方誤差比較

                      Table  1  Mean square error comparison

                      不同切片層的圖像 配準前 文獻 [7] 文獻 [13] 本文方法
                      I 0.1795 0.1467 0.1417 0.0013
                      II 0.1524 0.1232 0.0938 0.0013
                      III 0.0868 0.0695 0.0662 0.0012
                      下載: 導出CSV

                      表  2  Dice ratio比較

                      Table  2  Dice ratio comparison

                      不同切片層的圖像 配準前 文獻 [7] 文獻 [13] 本文方法
                      I 0.3925 0.5107 0.5859 0.9793
                      II 0.4702 0.5874 0.5836 0.9734
                      III 0.4683 0.5124 0.5874 0.9808
                      下載: 導出CSV

                      表  3  冠狀面配準精度對比

                      Table  3  Comparison of registration accuracy of coronal plane

                      MSE Dice ratio
                      配準前 0.0798 0.4392
                      文獻 [7] 0.0605 0.6345
                      文獻 [13] 0.0412 0.7498
                      本文算法 0.0061 0.9461
                      下載: 導出CSV

                      表  4  矢狀面配準精度對比

                      Table  4  Comparison of registration accuracy of sagittal plane

                      MSE Dice ratio
                      配準前 0.0910 0.4935
                      文獻 [7] 0.0598 0.6972
                      文獻 [13] 0.0249 0.7854
                      本文算法 0.0093 0.9278
                      下載: 導出CSV

                      表  5  橫切面配準精度對比

                      Table  5  Comparison of registration accuracy of transverse plane

                      MSE Dice ratio
                      配準前 0.0689 0.5214
                      文獻 [7] 0.0497 0.6743
                      文獻 [13] 0.0163 0.8057
                      本文算法 0.0049 0.9649
                      下載: 導出CSV

                      表  6  不同算法的時間對比(s)

                      Table  6  Time comparison of two methods (s)

                      不同切片層的圖像 文獻 [13] 的方法 (不同的階次) 本文方法
                      $ \alpha $= 0.1 $ \alpha $= 0.2 $ \alpha $= 0.3 $ \alpha $= 0.4 $ \alpha $= 0.5 $ \alpha $= 0.6 $ \alpha $= 0.7 $ \alpha $= 0.8 $ \alpha $= 0.9 總計時間
                      I 3.21 3.14 2.98 2.76 3.03 2.89 2.92 2.67 2.58 26.18 17.69
                      II 3.72 3.65 3.37 3.54 3.68 3.03 3.29 3.21 3.16 30.65 19.08
                      III 4.64 4.61 4.53 4.57 4.65 4.06 4.52 4.18 4.49 40.25 26.83
                      下載: 導出CSV

                      表  7  兩種策略的時間對比(s)

                      Table  7  Time comparison of two strategies (s)

                      圖像 不采用多分辨率 采用多分辨率
                      I 30.25 17.69
                      II 28.04 19.08
                      III 40.63 26.83
                      下載: 導出CSV
                      360彩票
                    • [1] 申艷平. 醫學圖像配準技術. 中國醫學物理雜志, 2013, 30(1): 3385?3389

                      Shen Yan-Ping. Review of image registration methods for medical images. Chinese Journal of Medical Physics, 2013, 30(1): 3385?3389
                      [2] Thirion J P. Image matching as a diffusion process: An analogy with Maxwell's demons. Medical Image Analysis, 1998, 2(3): 243?260 doi: 10.1016/S1361-8415(98)80022-4
                      [3] Wang H, Dong L, O′Daniel J, Mohan R, Garden A S, Ang K K, et al. Validation of an accelerated “demons” algorithm fordeformable image registration in radiation therapy. Physics in Medicine and Biology, 2005, 50(12): 2887?2905 doi: 10.1088/0031-9155/50/12/011
                      [4] Ying S H, Li D, Xiao B, Peng Y X, Do S Y, Xu M F. Nonlinear image registration with bidirectional metric and reciprocal regularization. PloS One, 2017, 12(2): e0172432 doi: 10.1371/journal.pone.0172432
                      [5] Du S Y, Guo Y R, Sanroma G, Ni D, Wu G R, Shen D G. Building dynamic population graph for accurate correspondence detection. Medical Image Analysis, 2015, 26(1): 256?267 doi: 10.1016/j.media.2015.10.001
                      [6] Du S Y, Zhang C J, Wu Z Z, Liu J, Xue J R. Robust isotropic sffigcaling ICP algorithm with bidirectional distance and bounded rotation angle. Neurocomputing, 2016, 215: 160?168
                      [7] Vercauteren T, Pennec X, Perchant A, Ayache N. Symmetric log-domain diffeomorphic registration: A demons-based approach. Med Image Comput Comput Assist Interv, 2008, 11(Pt1): 754?761
                      [8] 郝培博, 陳震, 江少鋒, 汪洋. 基于Demons算法的多模態醫學圖像非剛性配準研究. 生物醫學工程學雜志, 2014, 1: 161?165

                      Hao P B, Chen Z, Jiang S F, Wang Y. Research on non-rigid registration of multi-modal medical image based on demons algorithm. Journal of Biomedical Engineering, 2014, 1: 161?165
                      [9] Lu X Q, Yu H F, Zhao Y, Hou H, Li Y H. Three-dimensional lung medical image registration based on improved demons algorithm. Optik–International Journal for Light and Electron Optics, 2015, 127(4): 1893?1899
                      [10] 閆德勤, 劉彩鳳, 劉勝藍, 劉德山. 大形變微分同胚圖像配準快速算法. 自動化學報, 2015, 41(8): 1461?1470

                      Yan De-Qin, Liu Cai-Feng, Liu Sheng-Lan, Liu De-Shan. A fast image registration algorithm for diffeomorphic image with large deformation. Acta Automatica Sinica, 2015, 41(8): 1461?1470
                      [11] 薛鵬, 楊佩, 曹祝樓, 賈大宇, 董恩清. 基于平衡系數的 Active demons非剛性配準算法. 自動化學報, 2016, 42(9): 1389?1400

                      Xue Peng, Yang Pei, Cao Zhu-Lou, Jia Da-Yu, Dong En-Qing. Active demons non-rigid registration algorithm based on balance coefficient. Acta Automatica Sinica, 2016, 42(9): 1389?1400
                      [12] 路玉昆, 鞏貫忠, 虞剛, 李登旺, 尹勇. 基于微分同胚Demons形變配準算法獲取肺通氣圖. 中國醫學物理學雜志, 2017, 34(7): 666?670

                      Lu Y K, Gong G Z, Yu G, Li D W, Yin Y. Diffeomorphic demons registration algorithm for obtaining lung ventilation image. Chinese Journal of Medical Physics, 2017, 34(7): 666?670
                      [13] 張桂梅, 曹紅洋, 陳陽泉, 劉建新. 基于分數階梯度驅動的主動Demons算法研究. 電子學報, 2016, 44(12): 2834?2841 doi: 10.3969/j.issn.0372-2112.2016.12.004

                      Zhang Gui-Mei, Cao Hong-Yang, Chen Yang-Quan, Liu Jian-Xin. Research on active demons based on fractional differentiation gradient driving. Acta Eletronica Sinica, 2016, 44(12): 2834?2841 doi: 10.3969/j.issn.0372-2112.2016.12.004
                      [14] Hu Y P, Modat M, Gibson E, Ghavami N, Bonmati E, Moore C M, Emberton M, Noble J A, Barratt D C, Vercauteren T. Label-driven weakly-supervised learning for multimodal deformable image registration. In: Proceedings of the 15th IEEE International Symposium on Biomedical Imaging, 2018. 1070–1074
                      [15] Balakrishnan G, Zhao A, Sabuncu M R, Guttag J, Dalca A V. An unsupervised learning model for deformable medical image registration. In: Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, UT, USA: IEEE, 2018. 235–242
                      [16] Pu Y F, Siarry P, Zhou J L, Zhang N. A fractional partial differential equation based multiscale denoising model for texture image. Mathematical Methods in the Applied Sciences, 2014, 37(12): 1784?1806 doi: 10.1002/mma.2935
                      [17] Chen D L, Chen Y Q, Xue D Y. Fractional-order total variation image denoising based on proximity algorithm. Applied Mathematics and Computation, 2015, 257: 537?545
                      [18] Mathieu B, Melchior P, Oustaloup A, Ceyral C. Fractional differentiation for edge detection. Signal Processing, 2003, 83(11): 2421?2432 doi: 10.1016/S0165-1684(03)00194-4
                      [19] Ren Z M. Adaptive active contour model driven by fractional order fitting energy. Signal Processing, 2015, 117: 138?150
                      [20] 張桂梅, 徐繼元, 劉建新. 一種新的基于自適應分數階的活動輪廓模型. 計算機研究與發展, 2017, 54(5): 1045?1056 doi: 10.7544/issn1000-1239.2017.20160301

                      Zhang Gui-Mei, Xu Ji-Yuan, Liu Jian-Xin. A new active contour model based on adaptive fractional order. Journal of Computer Research and Development, 2017, 54(5): 1045?1056 doi: 10.7544/issn1000-1239.2017.20160301
                      [21] Yu J M, Tan L J, Zhou S B, Wang L P, Siddique M A. Image denoising algorithm based on entropy and adaptive fractional order calculus operator. IEEE Access, 2017, 5: 12275?12285 doi: 10.1109/ACCESS.2017.2718558
                      [22] 張桂梅, 郭黎娟, 熊邦書, 儲珺. 基于多分辨率和自適應分數階的Active Demons算法. 計算機研究與發展, 2018, 55(12): 2753?2763 doi: 10.7544/issn1000-1239.2018.20170523

                      Zhang Gui-Mei, Guo Li-Juan, Xiong Bang-Shu, Chu Jun. Active demons algorithm based on multi-resolution and adaptive fractional differential. Journal of Computer Research and Development, 2018, 55(12): 2753?2763 doi: 10.7544/issn1000-1239.2018.20170523
                      [23] Du S Y, Xu G L, Zhang S R, Zhang X T, Gao Y, Chen B D. Robust rigid registration algorithm based on pointwise correspondence and correntropy. Pattern Recognition Letters, 2020, 132: 91?98
                    • 加載中
                    圖(5) / 表(7)
                    計量
                    • 文章訪問數:  234
                    • HTML全文瀏覽量:  71
                    • PDF下載量:  84
                    • 被引次數: 0
                    出版歷程
                    • 收稿日期:  2019-01-27
                    • 錄用日期:  2019-09-24
                    • 網絡出版日期:  2020-09-28
                    • 刊出日期:  2020-09-28

                    目錄

                      /

                      返回文章
                      返回