基于神经网络的脑组织图像分割
|
摘要
对磁共振脑图像中的脑组织进行分割是对脑组织进行定量分析的关键步骤。MR脑图像中不同的脑组织(灰质、白质、脑脊液)互相混迭在一起而没有明显的边界,MR成像过程中形成的伪影比其他医学成像方式多得多(如化学位移伪影、运动伪影、磁敏感性伪影等),磁场的不均匀性也会造成成像的偏差。这些因素使得要精确分割出不同的脑组织非常困难。
本文针对脑组织图像分割问题进行了研究。首先将神经网络应用于磁共振图像分割技术研究中,采用局部刺激全局抑制神经元振荡网络模型对脑组织图像进行分割,结果同传统的阈值分割结果进行对比。对脑组织图像进行预分割工作,目的是剔除非脑组织,保留大脑结构。提出改进的模糊自组织神经网络模型(IFKCN),对原有算法进行两点改进,进行模型性能分析。应用模型对预分割后的脑组织图像进行细分割工作,目的是分割出白质,灰质、脑脊液等脑局部组织,在分割的过程中对网络的关键参数进行研究。
本课题的研究工作和创新主要包括以下几点:
(1)对磁共振图像的神经网络算法进行研究,采用局部刺激全局抑制神经元振荡网络模型进行脑组织图像分割,通过与传统的阈值分割结果进行对比证明其算法优越性。
(2)对颅脑图像进行预分割工作。提出基于区域生长法和基于边界跟踪的大脑边界线提取方法,通过将结果进行比对证明区域跟踪边界提取算法更具稳健型。
(3)对原模糊神经网络学习算法进行两点改进:隶属度更新函数的调整,引入调整系数。提出一种改进的模糊Kohonen自组织神经网络模型(IFKCN)。通过实验进行模型性能分析证明其优越性。
(4)应用改进后模型对脑组织图像进行细分割工作,给出改进前后网络迭代次数和收敛速度对比证明其优越性。对网络关键参数进行研究,提出兼顾单个像素以及邻域像素信息的输入矢量设计,引入聚类有效性函数评估进行聚类个数选择,给出实验结果。
Brain tissue segmentation of MR brain images is the key step for brain quantitative analysis. Compared with other medical image segmentation method, MR brain image segmentation is more challenging. This is mainly because, first, there is no clear border between different brain tissues (gray matter, white matter, cerebrospinal fluid), beside that, MR imaging process has more formation of artifacts than other medical imaging modalities (such as chemical displacement artifact, motion artifact, magnetic susceptibility artifacts, etc.), magnetic field inhomogeneity of the deviation may also affect the imaging. These factors make accurate segmentation of different brain tissue difficult. That is why a growing number of scholars get into the study of MR brain image segmentation.
Fist made research on segmentation of magnetic resonance technology using neural network, and made Application of LEGION(Locally Excitatory Globally Inhibitory Neuronal Oscillator Network)to segment images of brain tissue. The results were compared with the traditional threshold segmentation. The brain images of pre-partition aims to exclude non-brain tissue and retain the structure of the brain. Then in this thesis, with two original algorithm improvements, the model performance analysis is made to the improved fuzzy self-organizing neural network model (IFKCN). On the basis of model, brain issue segmentation aims to part the white matter (WM)、gray matter (GM)、cerebrospinal fluid (CSF). In the process of partition, the key parameters of the network were studied.
The main contributions of this thesis are mainly lie in the following aspects:
1. Made research on MR brain image segmentation based on neural network. The local stimulate and overall reject oscillatory model was adopt to segment the brain issue, the results were compared with the traditional threshold segmentation.
2. Pre-segmentation work. This paper proposed two brain boundary line extraction methods based on region growing and boundary tracking method. A comparison is mad between the two ones.
3. Two improvements were made to the original fuzzy neural network: updated membership of adjustment and the introduced the adjustment factor. This paper proposed an improved fuzzy kohonen self-organizing neural network model (IFKCN). Performance analysis was made on the model.
4. Applied the improved model for brain segmentation. Results were compared with the original one. In the segmentation process, research was made on two key paraments: the input pixel vector design and introduced the clustering validity function to assess.
本文针对脑组织图像分割问题进行了研究。首先将神经网络应用于磁共振图像分割技术研究中,采用局部刺激全局抑制神经元振荡网络模型对脑组织图像进行分割,结果同传统的阈值分割结果进行对比。对脑组织图像进行预分割工作,目的是剔除非脑组织,保留大脑结构。提出改进的模糊自组织神经网络模型(IFKCN),对原有算法进行两点改进,进行模型性能分析。应用模型对预分割后的脑组织图像进行细分割工作,目的是分割出白质,灰质、脑脊液等脑局部组织,在分割的过程中对网络的关键参数进行研究。
本课题的研究工作和创新主要包括以下几点:
(1)对磁共振图像的神经网络算法进行研究,采用局部刺激全局抑制神经元振荡网络模型进行脑组织图像分割,通过与传统的阈值分割结果进行对比证明其算法优越性。
(2)对颅脑图像进行预分割工作。提出基于区域生长法和基于边界跟踪的大脑边界线提取方法,通过将结果进行比对证明区域跟踪边界提取算法更具稳健型。
(3)对原模糊神经网络学习算法进行两点改进:隶属度更新函数的调整,引入调整系数。提出一种改进的模糊Kohonen自组织神经网络模型(IFKCN)。通过实验进行模型性能分析证明其优越性。
(4)应用改进后模型对脑组织图像进行细分割工作,给出改进前后网络迭代次数和收敛速度对比证明其优越性。对网络关键参数进行研究,提出兼顾单个像素以及邻域像素信息的输入矢量设计,引入聚类有效性函数评估进行聚类个数选择,给出实验结果。
Brain tissue segmentation of MR brain images is the key step for brain quantitative analysis. Compared with other medical image segmentation method, MR brain image segmentation is more challenging. This is mainly because, first, there is no clear border between different brain tissues (gray matter, white matter, cerebrospinal fluid), beside that, MR imaging process has more formation of artifacts than other medical imaging modalities (such as chemical displacement artifact, motion artifact, magnetic susceptibility artifacts, etc.), magnetic field inhomogeneity of the deviation may also affect the imaging. These factors make accurate segmentation of different brain tissue difficult. That is why a growing number of scholars get into the study of MR brain image segmentation.
Fist made research on segmentation of magnetic resonance technology using neural network, and made Application of LEGION(Locally Excitatory Globally Inhibitory Neuronal Oscillator Network)to segment images of brain tissue. The results were compared with the traditional threshold segmentation. The brain images of pre-partition aims to exclude non-brain tissue and retain the structure of the brain. Then in this thesis, with two original algorithm improvements, the model performance analysis is made to the improved fuzzy self-organizing neural network model (IFKCN). On the basis of model, brain issue segmentation aims to part the white matter (WM)、gray matter (GM)、cerebrospinal fluid (CSF). In the process of partition, the key parameters of the network were studied.
The main contributions of this thesis are mainly lie in the following aspects:
1. Made research on MR brain image segmentation based on neural network. The local stimulate and overall reject oscillatory model was adopt to segment the brain issue, the results were compared with the traditional threshold segmentation.
2. Pre-segmentation work. This paper proposed two brain boundary line extraction methods based on region growing and boundary tracking method. A comparison is mad between the two ones.
3. Two improvements were made to the original fuzzy neural network: updated membership of adjustment and the introduced the adjustment factor. This paper proposed an improved fuzzy kohonen self-organizing neural network model (IFKCN). Performance analysis was made on the model.
4. Applied the improved model for brain segmentation. Results were compared with the original one. In the segmentation process, research was made on two key paraments: the input pixel vector design and introduced the clustering validity function to assess.
引文
[1]Jagath C.Rajapakse.Statistical Approach to Segmentation of Single-Channel Cerebral MR Images[J].IEEE Trans.Medical Imaging,1997,16(2):176-186.
[2]J.C.Bezdek,et al.Review of MR image segmentation techniques using pattern recognition[J].Med.Phys,1993,20(4):1033-1048.
[3]S.Cagnoni,G.Coppini,M.Rucci,D.Caramella,G.Valli.Neural network segmentation of magnetic resonance spin echo image of the brain[J].Biomed.Eng,1993,15:355-362.
[4]S.C.Amartur,D.Piraino,and Y.Takefuji.Optimization Neural networks for segmentation of magnetic resonance images[J].IEEE Trans.Med,Imaging,1992,11:215-220.
[5]Arvid Lundervold & Geir Storvik.Segmengtation of Brain Parenchyma and Cerebrospinal Fluid in Multispectral Magnetic Resonance Images[J].IEEE Trans.Medical Imaging,1995,14(2):339-349.
[6]Tolias Yannis A,et al.A Fuzzy Vessel Tracking Algorithm for Retinal Images Based on Fuzzy Clustering[J],IEEE Trans.Medical Imaging,1998,17(2):263-273.
[7]赵喜平.磁共振成像系统的原理及其应用[M].科学出版社,2000.
[8]McCulloch W S,Pitts W.A Logical Calculus of the Ideas Immarent in Nervous Activity[J].Bulletin of Mathematical Biophysics,1943,(5):115-133.
[9]Hebb D O.The Organization of Behavior[M].New York:Wiley,1949.
[10]袁曾任.人工神经元网络及其应用[M].清华大学出版社,1999,10.
[11]常铜,张建龙,李丽雅.神经网络及其应用[J].航空计算技术,1999(4):50-53.
[12]文传源.人工神经网络及其应用[J].计算机仿真,1997(2):10-12.
[13]聂生东,章鲁,顾顺德,陈瑛.磁共振图像的分割[J].国外医学生物工程分册,1999,22(6):349-355.
[14]L.O.Hall,A.M.Bensain,L.P.Clarke,R.P.Velthuizen,M.Silbiger,J.C.Bezdek.A comparison of neural networks and fuzzy clustering techniques in segmentation magnetic resonance of the brain [J].IEEE Trans.On Neural Networks.1992,3:672-683.
[15]Ulrich Raff et.Quantitation of grey matter,white matter,and cerebrospinal fluid from spin-echo magnetic resonance images using an artificial neural network technique [J].Med.Phys.1994,21:1933-1942.
[16]B.Kosko.Fuzzy associative memories.In:Kondel(ed).Fuzzy Expert System Reanding [M].MA:Addison Weley,1987.
[17]S.C.Amartur,D.Piraino,and Y.Takefuji.Optimization Neural networks for the segmentation of magnetic resonance images [J].IEEE Trans.Med Imaging,1992,11:215-220.
[18]Zeyun Yu,Chandrajit Bajaj.Image segmentation using gradient vector diffusion and region merging[C].Quebec,Canada ".Proceedings of the 16th International Conference Pattern Recognition,2002:941-944.
[19]K.S.Cheng,J.S.Lin,and C.WMao.The application of competitive Hopfield neural network to medical image segmentation.IEEE Trans.Med.Imaging,1996,15:560-567.
[20]S,Haring,M.A.Viergever and J.N.Kok.Kohonen networks from ultiscale image segmentation [J].Image and Vision Computing,1994,12(6):339-344.
[21]A.P.Dhawan,L.Arata.Segmentation of medical images through competitive learning [J].Comput.Meth.Prog.Biomed,1993,40:203-215.
[22]ZHANG Hong-mei,YUAN Ze-jian,CAI Zhong-min,et al.Segmentation of MRI using hierarchical markov random field[J].Journal of Software,2002,13(9):1779-1808.
[23]S.Cagnoni,G.Coppini,M.Rucci,D.Caramel la,G.Vail i.Neural network segmentation of magnetic resonance spin echo image of the brain [J].J.Biomed ,End,1993,15:355-362.
[24]A.J.Worth,D.N.Kennedy.Segmentation of magnetic resonance brain images using analogue constraint satisfaction neural networks [J].Image and Vision Computing,1994,12:345-354.
[25]W.C.Lin,E.CC.K.Tsao and C.T.Chen.Constraint sat if action neural networks for image segmentation[J].Patern Recognition,1992,25:679-693.
[26]WSnyder,A.Logenthiram,P.Santago,K,Link,G.Bilbro,S.Rajala.Segmentation brain images using mean field annealing.In inform.Process.Med.Imag.Proc.12~(th)Int.Conf.IPMI.Wye,UK,1991:218-226.
[27]N.Shareef D.L.Wang,and R.Yagel.Segmentation of medical images using LEGION.IEEE Trans.Med,Imaging,1999,18:74-91.
[28]顾顺德,聂生东,陈瑛,章鲁.基于K-最近邻规则的磁共振颅脑图像分割算法的应用研究[J].上海医科大学学报,2000,Mar,27(2).
[29]陈方昕.基于区域生长法的图像分割技术[J].信息科技,2008年第15期.
[30]Rafael C.Gonzalez,Richard E.Woods.Digital Image Processing Second Edition [M].电子工业出版社,2004.
[31]文传源.人工神经网络及其应用[J].计算机仿真,1997(2):10-12.
[32]Kohonen T.Self-organization and Associative Memory.3rdEdn,Springer,Berlin (1989).
[33]Huntsberger T,Ajjimarangsee P.Parallel Self-organization feature maps for unsupervised Pattern recognition.Int.j.Gen Syst.,1989(16):357-372.
[34]Kohonen T,Mahisara K,Saramaki T.Phonotopic Maps-Insightful Representation of Phonological Features for Speech Recognition.Proc.7th Int'l Conf.on Pattern Recognition Montreal,Canada,1984:82-185.
[35]Whittington G,Spracklen C T.The Application of Neural Networks to Tactical and Sensor Data Fusion Problems.First IEE international conference on Artificial neural networks,London,1989:342-345.
[36]郁文贤,胡卫东等.一种新的弱目标自组织模糊分类方法[J].电子学报,1996(6):67-71.
[37]Yu Xiaohan,Juha Yla-jaaski.A General Image Segmentation Method Combining Region Growing and Edge Dtection[J].北方交通大学学报,第21卷第1期,1997年2月.
[38]罗述谦.医学图像处理与分析[M].科学出版社,2003.
[39]L A.Zadeh.Fuzzy sets[J].Information and Control,1965,8:338-353.
[40]Whittington G,Spracklen C T.The Application of Neural Networks to Tactical and Sensor Data Fusion Problems,First IEE international conference on Artificial neural networks,London,1989:342-345.
[41]蔡元龙.模式识别[M].西安电子科技大学出版社,1990.
[42]Xuanl,Xie LS.BeniG.A Vitality measure for fuzzy istering[J].IEEE Trans on Pattern Analysis and Machine Intelligence,1992,13(8):841-845.
[43]Gindy NN,Z Ratchev TM,Case K.Component grouping for CT applications-a fuzzy clustering approach with validity measure[J].International Journal of Production Research,1995,33(9):2493.
[2]J.C.Bezdek,et al.Review of MR image segmentation techniques using pattern recognition[J].Med.Phys,1993,20(4):1033-1048.
[3]S.Cagnoni,G.Coppini,M.Rucci,D.Caramella,G.Valli.Neural network segmentation of magnetic resonance spin echo image of the brain[J].Biomed.Eng,1993,15:355-362.
[4]S.C.Amartur,D.Piraino,and Y.Takefuji.Optimization Neural networks for segmentation of magnetic resonance images[J].IEEE Trans.Med,Imaging,1992,11:215-220.
[5]Arvid Lundervold & Geir Storvik.Segmengtation of Brain Parenchyma and Cerebrospinal Fluid in Multispectral Magnetic Resonance Images[J].IEEE Trans.Medical Imaging,1995,14(2):339-349.
[6]Tolias Yannis A,et al.A Fuzzy Vessel Tracking Algorithm for Retinal Images Based on Fuzzy Clustering[J],IEEE Trans.Medical Imaging,1998,17(2):263-273.
[7]赵喜平.磁共振成像系统的原理及其应用[M].科学出版社,2000.
[8]McCulloch W S,Pitts W.A Logical Calculus of the Ideas Immarent in Nervous Activity[J].Bulletin of Mathematical Biophysics,1943,(5):115-133.
[9]Hebb D O.The Organization of Behavior[M].New York:Wiley,1949.
[10]袁曾任.人工神经元网络及其应用[M].清华大学出版社,1999,10.
[11]常铜,张建龙,李丽雅.神经网络及其应用[J].航空计算技术,1999(4):50-53.
[12]文传源.人工神经网络及其应用[J].计算机仿真,1997(2):10-12.
[13]聂生东,章鲁,顾顺德,陈瑛.磁共振图像的分割[J].国外医学生物工程分册,1999,22(6):349-355.
[14]L.O.Hall,A.M.Bensain,L.P.Clarke,R.P.Velthuizen,M.Silbiger,J.C.Bezdek.A comparison of neural networks and fuzzy clustering techniques in segmentation magnetic resonance of the brain [J].IEEE Trans.On Neural Networks.1992,3:672-683.
[15]Ulrich Raff et.Quantitation of grey matter,white matter,and cerebrospinal fluid from spin-echo magnetic resonance images using an artificial neural network technique [J].Med.Phys.1994,21:1933-1942.
[16]B.Kosko.Fuzzy associative memories.In:Kondel(ed).Fuzzy Expert System Reanding [M].MA:Addison Weley,1987.
[17]S.C.Amartur,D.Piraino,and Y.Takefuji.Optimization Neural networks for the segmentation of magnetic resonance images [J].IEEE Trans.Med Imaging,1992,11:215-220.
[18]Zeyun Yu,Chandrajit Bajaj.Image segmentation using gradient vector diffusion and region merging[C].Quebec,Canada ".Proceedings of the 16th International Conference Pattern Recognition,2002:941-944.
[19]K.S.Cheng,J.S.Lin,and C.WMao.The application of competitive Hopfield neural network to medical image segmentation.IEEE Trans.Med.Imaging,1996,15:560-567.
[20]S,Haring,M.A.Viergever and J.N.Kok.Kohonen networks from ultiscale image segmentation [J].Image and Vision Computing,1994,12(6):339-344.
[21]A.P.Dhawan,L.Arata.Segmentation of medical images through competitive learning [J].Comput.Meth.Prog.Biomed,1993,40:203-215.
[22]ZHANG Hong-mei,YUAN Ze-jian,CAI Zhong-min,et al.Segmentation of MRI using hierarchical markov random field[J].Journal of Software,2002,13(9):1779-1808.
[23]S.Cagnoni,G.Coppini,M.Rucci,D.Caramel la,G.Vail i.Neural network segmentation of magnetic resonance spin echo image of the brain [J].J.Biomed ,End,1993,15:355-362.
[24]A.J.Worth,D.N.Kennedy.Segmentation of magnetic resonance brain images using analogue constraint satisfaction neural networks [J].Image and Vision Computing,1994,12:345-354.
[25]W.C.Lin,E.CC.K.Tsao and C.T.Chen.Constraint sat if action neural networks for image segmentation[J].Patern Recognition,1992,25:679-693.
[26]WSnyder,A.Logenthiram,P.Santago,K,Link,G.Bilbro,S.Rajala.Segmentation brain images using mean field annealing.In inform.Process.Med.Imag.Proc.12~(th)Int.Conf.IPMI.Wye,UK,1991:218-226.
[27]N.Shareef D.L.Wang,and R.Yagel.Segmentation of medical images using LEGION.IEEE Trans.Med,Imaging,1999,18:74-91.
[28]顾顺德,聂生东,陈瑛,章鲁.基于K-最近邻规则的磁共振颅脑图像分割算法的应用研究[J].上海医科大学学报,2000,Mar,27(2).
[29]陈方昕.基于区域生长法的图像分割技术[J].信息科技,2008年第15期.
[30]Rafael C.Gonzalez,Richard E.Woods.Digital Image Processing Second Edition [M].电子工业出版社,2004.
[31]文传源.人工神经网络及其应用[J].计算机仿真,1997(2):10-12.
[32]Kohonen T.Self-organization and Associative Memory.3rdEdn,Springer,Berlin (1989).
[33]Huntsberger T,Ajjimarangsee P.Parallel Self-organization feature maps for unsupervised Pattern recognition.Int.j.Gen Syst.,1989(16):357-372.
[34]Kohonen T,Mahisara K,Saramaki T.Phonotopic Maps-Insightful Representation of Phonological Features for Speech Recognition.Proc.7th Int'l Conf.on Pattern Recognition Montreal,Canada,1984:82-185.
[35]Whittington G,Spracklen C T.The Application of Neural Networks to Tactical and Sensor Data Fusion Problems.First IEE international conference on Artificial neural networks,London,1989:342-345.
[36]郁文贤,胡卫东等.一种新的弱目标自组织模糊分类方法[J].电子学报,1996(6):67-71.
[37]Yu Xiaohan,Juha Yla-jaaski.A General Image Segmentation Method Combining Region Growing and Edge Dtection[J].北方交通大学学报,第21卷第1期,1997年2月.
[38]罗述谦.医学图像处理与分析[M].科学出版社,2003.
[39]L A.Zadeh.Fuzzy sets[J].Information and Control,1965,8:338-353.
[40]Whittington G,Spracklen C T.The Application of Neural Networks to Tactical and Sensor Data Fusion Problems,First IEE international conference on Artificial neural networks,London,1989:342-345.
[41]蔡元龙.模式识别[M].西安电子科技大学出版社,1990.
[42]Xuanl,Xie LS.BeniG.A Vitality measure for fuzzy istering[J].IEEE Trans on Pattern Analysis and Machine Intelligence,1992,13(8):841-845.
[43]Gindy NN,Z Ratchev TM,Case K.Component grouping for CT applications-a fuzzy clustering approach with validity measure[J].International Journal of Production Research,1995,33(9):2493.