(34.204.201.220) 您好!臺灣時間:2021/04/19 17:15
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:鄭綉瑩
研究生(外文):Hsiu-YingCheng
論文名稱:利用高斯混合模型自動化分割核磁共振影像中之多發性硬化症病變
論文名稱(外文):Automatic segmentation of Multiple Sclerosis Lesion in MRIusing Gaussian Mixture Model
指導教授:吳明龍趙梓程
指導教授(外文):Ming-Long WuTzu-Cheng Chao
學位類別:碩士
校院名稱:國立成功大學
系所名稱:醫學資訊研究所
學門:醫藥衛生學門
學類:醫學技術及檢驗學類
論文種類:學術論文
論文出版年:2018
畢業學年度:106
語文別:英文
論文頁數:67
中文關鍵詞:自動化分割多發性硬化症多發性硬化症自動分割白質病變磁振造影高斯混合模型
外文關鍵詞:Automatic segmentation of Multiple Sclerosis lesionsMultiple SclerosisAutomatic segmentationWhite matter lesionMagnetic Resonance ImagingGaussian Mixture Model
相關次數:
  • 被引用被引用:0
  • 點閱點閱:56
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
為了降低手動分割多發性硬化症的病變所造成的問題,如耗時、評估者之間的差異,所以越來越多人提出多發性硬化症的病變的自動分割法。本論文針對多發性硬化症的白質病變進行自動分割,使用了三種不同的磁振造影序列(T1-w, T2-w, FLAIR 影像)。我們主要使用高斯混合模型來評估正常組織的高斯分佈的參數,其中會先評估出可能是離群值的部分並扣除掉,再評估出正常組織的參數,由於多發性硬化症的白質病變在T2-w 和FLAIR 影像中屬於高強度,所以將評估出來的離群值對應到這兩種影像上來找到高強度的地方,也就是多發性硬化症白質病變的位置,最後執行後處裡,將偽陽性排除。將最後得到的病變計算出體積,可以幫助醫師長期監督病變的變化。
Manual lesion segmentation of Multiple Sclerosis (MS) lesions is time consuming and has an intra- and interrater variability. To resolve these issues, there are several auto-segmentation methods. This thesis implements an automatic MS lesion segmentation method targeting white matter lesions (WML), uses three different MRI sequences (T1-w, T2-w, FLAIR image) as the input of our algorithm. We use Gaussian Mixture Model (GMM) to estimate the parameters of Gaussian distribution of NABT (normal appearing brain tissue) and we will remove the outlier candidate data during this process to ensure correctness. MS WML has hyper-intensity on T2-w and FLAIR images, so we find the hyper-intensity pixels on these two kinds of images using the outlier information estimate from the previous step. In the post-processing, we remove false positive using rule-based methods and evaluate the volumes of lesions to help doctors supervise the progress of lesion in long term.
摘要. . . . . . . . . . . . . . . . . . . . . . . . . . i
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . ii
誌謝. . . . . . . . . . . . . . . . . . . . . . . . . . iii
Table of Contents . . . . . . . . . . . . . . . . . . . . . . . . . iv
Chapter 1 Introduction . . . . . . . . . . . . . . . . . . . . . . .1
1.1 What is Multiple Sclerosis . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 MRI in brain disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Segmentation of Multiple Sclerosis lesions . . . . . . . . . . . . . . . . . . 2
1.4 Challenges in detection of white matter lesions . . . . . . . . . . . . . . . 4
Chapter 2 Materials and Methods . . . . . . . . . . . . . . . . . . 5
2.1 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Pre-processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2.1 Skull stripping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2.2 Intensity inhomogeneity correction . . . . . . . . . . . . . . . . . . 8
2.2.3 Registration of three sequences . . . . . . . . . . . . . . . . . . . . 8
2.3 Estimation of normal appearing brain tissue (NABT) parameter . . . . . . . 9
2.3.1 K-means++ clustering for initialization of parameters in GMM . . . 9
2.3.2 Gaussian Mixture Model for finding parameters of Normal Appearing
Brain Tissue (NABT) . . . . . . . . . . . . . . . . . . . . . . . . 11
2.3.3 Classification of NABT . . . . . . . . . . . . . . . . . . . . . . . . 14
2.4 Detection of lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.4.1 Mahalanobis Distance . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.4.2 Finding pixels with hyper-intensity . . . . . . . . . . . . . . . . . . 16
2.5 Post-processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.5.1 Filtering of lesion size . . . . . . . . . . . . . . . . . . . . . . . . 18
2.5.2 Filtering of neighbor information . . . . . . . . . . . . . . . . . . . 19
2.6 Evaluation of segmentation results . . . . . . . . . . . . . . . . . . . . . . 19
Chapter 3 Results. . . . . . . . . . . . . . . . . . . . . . . 22
3.1 Trimmed likelihood estimator (TLE) . . . . . . . . . . . . . . . . . . . . . 22
3.2 Classification of NABT . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.3 Mahalanobis distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.4 Automatic thresholding of hyper-intensity . . . . . . . . . . . . . . . . . . 28
3.5 Post-processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.5.1 Filtering of lesion size . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.5.2 Filtering by neighboring tissue type . . . . . . . . . . . . . . . . . . 30
3.6 Evaluation of segmentation results . . . . . . . . . . . . . . . . . . . . . . 33
3.6.1 Data from 2015 International Symposium on Biomedical Imagine . . 33
3.6.2 Data from National Cheng Kung University Hospital . . . . . . . . 50
3.7 Computation time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Chapter 4 Discussion . . . . . . . . . . . . . . . . . . . . . . .55
4.1 Stability of K-means . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.2 Trimmed likelihood estimator (TLE) . . . . . . . . . . . . . . . . . . . . . 55
4.3 Classification of NABT . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.4 Automatic thresholding of hyper-intensity . . . . . . . . . . . . . . . . . . 57
4.5 Post-processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.5.1 Filtering of neighbor information rule . . . . . . . . . . . . . . . . 58
4.6 Evaluation of segmentation results . . . . . . . . . . . . . . . . . . . . . . 58
Chapter 5 Future Work. . . . . . . . . . . . . . . . . . . . . . .64
References. . . . . . . . . . . . . . . . . . . . . . . . . . 65
[1] Benjamin KT Tsang, Richard Macdonell, et al. Multiple sclerosis: diagnosis, management
and prognosis. Australian family physician, 40(12):948, 2011.
[2] Samantha S Soldan, Gregory Wu, Clyde Markowitz, and Dennis L Kolson. Multiple
sclerosis and other demyelinating diseases. Neuroimmune Pharmacology, pages 239–
255, 2008.
[3] Daniel García-Lorenzo. Robust segmentation of focal lesions on multi-sequence MRI
in multiple sclerosis. PhD thesis, Université Rennes 1, 2010.
[4] Shannon J Kenealy, Margaret A Pericak-Vance, and Jonathan L Haines. The genetic
epidemiology of multiple sclerosis. Journal of neuroimmunology, 143(1):7–12, 2003.
[5] Massimo Filippi and MA Rocca. Mr imaging of gray matter involvement in multiple
sclerosis: implications for understanding disease pathophysiology and monitoring
treatment efficacy. American Journal of Neuroradiology, 31(7):1171–1177, 2010.
[6] Daniel García-Lorenzo, Simon Francis, Sridar Narayanan, Douglas L Arnold, and
D Louis Collins. Review of automatic segmentation methods of multiple sclerosis white
matter lesions on conventional magnetic resonance imaging. Medical image analysis,
17(1):1–18, 2013.
[7] Olga Ciccarelli, Peter A. Brex, Alan J. Thompson, and David H. Miller. Disability and
lesion load in ms: a reassessment with ms functional composite score and 3d fast flair.
Journal of Neurology, 249(1):18–24, Jan 2002.
[8] M Filippi, MA Horsfield, A Campi, S Mammi, C Pereira, and G Comi. Resolutiondependent
estimates of lesion volumes in magnetic resonance imaging studies of the
brain in multiple sclerosis. Annals of neurology, 38(5):749–754, 1995.
[9] Nancy L Sicotte, Rhonda R Voskuhl, Seth Bouvier, Rochelle Klutch, Mark S Cohen,
and John C Mazziotta. Comparison of multiple sclerosis lesions at 1.5 and 3.0 tesla.
Investigative radiology, 38(7):423–427, 2003.
[10] Yuri Y Boykov and M-P Jolly. Interactive graph cuts for optimal boundary & region
segmentation of objects in nd images. In Computer Vision, 2001. ICCV 2001. Proceedings.
Eighth IEEE International Conference on, volume 1, pages 105–112. IEEE,
2001.
[11] Claudia Lucchinetti, Wolfgang Bruck, Joseph Parisi, Bernd Scheithauer, Moses Rodriguez,
Hans Lassman, et al. Heterogeneity of multiple sclerosis lesions: implications
for the pathogenesis of demyelination. Annals of neurology, 47(6):707–717, 2000.
[12] 2015 longitudinal ms lesion segmentation challenge, 2015. http://www.iacl.jhu.
edu/index.php/MSChallenge.
[13] Mipav, 2016. https://mipav.cit.nih.gov/index.php.
[14] Marcello Alecci, Christopher M Collins, Michael B Smith, and Peter Jezzard. Radio
frequency magnetic field mapping of a 3 tesla birdcage coil: experimental and theoretical
dependence on sample properties. Magnetic resonance in medicine, 46(2):379–385,
2001.
[15] Nicholas J Tustison, Brian B Avants, Philip A Cook, Yuanjie Zheng, Alexander Egan,
Paul A Yushkevich, and James C Gee. N4itk: improved n3 bias correction. IEEE
transactions on medical imaging, 29(6):1310–1320, 2010.
[16] John G Sled, Alex P Zijdenbos, and Alan C Evans. A nonparametric method for automatic
correction of intensity nonuniformity in mri data. IEEE transactions on medical
imaging, 17(1):87–97, 1998.
[17] Mipav tool - optimized automatic registration 3d, 2016. https://mipav.cit.nih.
gov/pubwiki/index.php/Optimized_automatic_registration_3D.
[18] Daniel Garcia-Lorenzo, Sylvain Prima, Douglas L Arnold, D Louis Collins, and Christian
Barillot. Trimmed-likelihood estimation for focal lesions and tissue segmentation
in multisequence mri for multiple sclerosis. IEEE Transactions on Medical Imaging,
30(8):1455–1467, 2011.
[19] Alfiia Galimzianova, Žiga Špiclin, Boštjan Likar, and Franjo Pernuš. Robust mixtureparameter
estimation for unsupervised segmentation of brain mr images. In International
MICCAI Workshop on Medical Computer Vision, pages 84–94. Springer, 2013.
[20] Daniel García-Lorenzo, Sylvain Prima, Sean Patrick Morrissey, and Christian Barillot.
A robust expectation-maximization algorithm for multiple sclerosis lesion segmentation.
In MICCAI Workshop: 3D Segmentation in the Clinic: A Grand Challenge II, MS
lesion segmentation, page 277, 2008.
[21] David Arthur and Sergei Vassilvitskii. k-means++: The advantages of careful seeding.
In Proceedings of the eighteenth annual ACM-SIAM symposium on Discrete algorithms,
pages 1027–1035. Society for Industrial and Applied Mathematics, 2007.
[22] J. B. MacQueen. Some methods for classification and analysis of multivariate observations.
In L. M. Le Cam and J. Neyman, editors, Proc. of the fifth Berkeley Symposium
on Mathematical Statistics and Probability, volume 1, pages 281–297. University of
California Press, 1967.
[23] Peter J Rousseeuw and Christophe Croux. Alternatives to the median absolute deviation.
Journal of the American Statistical association, 88(424):1273–1283, 1993.
[24] Laure S Aït-Ali, Sylvain Prima, Pierre Hellier, Béatrice Carsin, Gilles Edan, and Christian
Barillot. Strem: a robust multidimensional parametric method to segment ms lesions
in mri. In International Conference on Medical Image Computing and Computer-
Assisted Intervention, pages 409–416. Springer, 2005.
[25] Miin-Shen Yang, Chien-Yo Lai, and Chih-Ying Lin. A robust em clustering algorithm
for gaussian mixture models. Pattern Recognition, 45(11):3950–3961, 2012.
[26] Neyko Neykov, Peter Filzmoser, R Dimova, and Plamen Neytchev. Robust fitting
of mixtures using the trimmed likelihood estimator. Computational Statistics & Data
Analysis, 52(1):299–308, 2007.
[27] Christopher M. Bishop. Pattern Recognition and Machine Learning. Springer, 2006.
[28] Implementation of em algorithm, 2016. http://www.cnblogs.com/549294286/p/
3273294.html.
[29] P. C. Mahalanobis. On the generalised distance in statistics. In Proceedings National
Institute of Science, India, pages 49–55, April 1936.
[30] Johanna Hardin and David M Rocke. The distribution of robust distances. Journal of
Computational and Graphical Statistics, 14(4):928–946, 2005.
[31] Eloy Roura, Arnau Oliver, Mariano Cabezas, Sergi Valverde, Deborah Pareto, Joan C
Vilanova, Lluís Ramió-Torrentà, Àlex Rovira, and Xavier Lladó. A toolbox for multiple
sclerosis lesion segmentation. Neuroradiology, 57(10):1031–1043, 2015.
[32] Onur Ganiler, Arnau Oliver, Yago Diez, Jordi Freixenet, Joan C Vilanova, Brigitte Beltran,
Lluís Ramió-Torrentà, Àlex Rovira, and Xavier Lladó. A subtraction pipeline for
automatic detection of new appearing multiple sclerosis lesions in longitudinal studies.
Neuroradiology, 56(5):363–374, 2014.
[33] Daniel García-Lorenzo, Jeremy Lecoeur, Douglas L Arnold, D Louis Collins, and
Christian Barillot. Multiple sclerosis lesion segmentation using an automatic multimodal
graph cuts. In International Conference on Medical Image Computing and
Computer-Assisted Intervention, pages 584–591. Springer, 2009.
[34] Alireza Karimian and Simin Jafari. A new method to segment the multiple sclerosis
lesions on brain magnetic resonance images. Journal of medical signals and sensors,
5(4):238, 2015.
[35] Jeremy Beaumont, Olivier Commowick, and Christian Barillot. Automatic multiple
sclerosis lesion segmentation from intensity-normalized multi-channel mri. In Proceedings
of the 1st MICCAI Challenge on Multiple Sclerosis Lesions Segmentation
Challenge Using a Data Management and Processing Infrastructure-MICCAI-MSSEG,
2016.
[36] Kelly H Zou, Simon K Warfield, Aditya Bharatha, Clare MC Tempany, Michael R
Kaus, Steven J Haker, William M Wells, Ferenc A Jolesz, and Ron Kikinis. Statistical
validation of image segmentation quality based on a spatial overlap index 1: Scientific
reports. Academic radiology, 11(2):178–189, 2004.
[37] Lee R Dice. Measures of the amount of ecologic association between species. Ecology,
26(3):297–302, 1945.
[38] Aaron Carass, Snehashis Roy, Amod Jog, Jennifer L Cuzzocreo, Elizabeth Magrath,
Adrian Gherman, Julia Button, James Nguyen, Ferran Prados, Carole H Sudre, et al.
Longitudinal multiple sclerosis lesion segmentation: Resource and challenge. NeuroImage,
148:77–102, 2017.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
系統版面圖檔 系統版面圖檔