Brain extraction based on locally linear representation-based classification

被引:31
作者
Huang, Meiyan [1 ]
Yang, Wei [1 ]
Jiang, Jun [1 ]
Wu, Yao [1 ]
Zhang, Yu [1 ]
Chen, Wufan [1 ]
Feng, Qianjin [1 ]
机构
[1] Southern Med Univ, Sch Biomed Engn, Guangzhou 510515, Guangdong, Peoples R China
基金
国家高技术研究发展计划(863计划); 美国国家卫生研究院;
关键词
Brain extraction; Locally Linear Representation-based; Classification; Label fusion; Local anchor embedding; SKULL STRIPPING PROBLEM; MRI DATA; AUTOMATIC SEGMENTATION; SPARSE REPRESENTATION; PROBABILISTIC ATLAS; META-ALGORITHM; IMAGES; MORPHOMETRY; MODEL;
D O I
10.1016/j.neuroimage.2014.01.059
中图分类号
Q189 [神经科学];
学科分类号
071006 ;
摘要
Brain extraction is an important procedure in brain image analysis. Although numerous brain extraction methods have been presented, enhancing brain extraction methods remains challenging because brain MRI images exhibit complex characteristics, such as anatomical variability and intensity differences across different sequences and scanners. To address this problem, we present a Locally Linear Representation-based Classification (LLRC) method for brain extraction. A novel classification framework is derived by introducing the locally linear representation to the classical classification model. Under this classification framework, a common label fusion approach can be considered as a special case and thoroughly interpreted. Locality is important to calculate fusion weights for LLRC; this factor is also considered to determine that Local Anchor Embedding is more applicable in solving locally linear coefficients compared with other linear representation approaches. Moreover, LLRC supplies a way to learn the optimal classification scores of the training samples in the dictionary to obtain accurate classification. The International Consortium for Brain Mapping and the Alzheimer's Disease Neuroimaging Initiative databases were used to build a training dataset containing 70 scans. To evaluate the proposed method, we used four publicly available datasets (IBSR1, IBSR2, LPBA40, and ADNI3T, with a total of 241 scans). Experimental results demonstrate that the proposed method outperforms the four common brain extraction methods (BET, BSE, GCUT, and ROBEX), and is comparable to the performance of BEaST, while being more accurate on some datasets compared with BEaST. (C) 2014 Elsevier Inc. All rights reserved.
引用
收藏
页码:322 / 339
页数:18
相关论文
共 39 条
[1]   Voxel-based morphometry - The methods [J].
Ashburner, J ;
Friston, KJ .
NEUROIMAGE, 2000, 11 (06) :805-821
[2]   Segmentation of brain 3D MR images using level sets and dense registration [J].
Baillard, C ;
Hellier, P ;
Barillot, C .
MEDICAL IMAGE ANALYSIS, 2001, 5 (03) :185-194
[3]   A review of image denoising algorithms, with a new one [J].
Buades, A ;
Coll, B ;
Morel, JM .
MULTISCALE MODELING & SIMULATION, 2005, 4 (02) :490-530
[4]   Simple paradigm for extra-cerebral tissue removal: Algorithm and analysis [J].
Carass, Aaron ;
Cuzzocreo, Jennifer ;
Wheeler, M. Bryan ;
Bazin, Pierre-Louis ;
Resnick, Susan M. ;
Prince, Jerry L. .
NEUROIMAGE, 2011, 56 (04) :1982-1992
[5]   Statistical morphological skull stripping of adult and infant MRI data [J].
Chiverton, John ;
Wells, Kevin ;
Lewis, Emma ;
Chen, Chao ;
Podda, Barbara ;
Johnson, Declan .
COMPUTERS IN BIOLOGY AND MEDICINE, 2007, 37 (03) :342-357
[6]   Cortical surface-based analysis - I. Segmentation and surface reconstruction [J].
Dale, AM ;
Fischl, B ;
Sereno, MI .
NEUROIMAGE, 1999, 9 (02) :179-194
[7]   BEaST: Brain extraction based on nonlocal segmentation technique [J].
Eskildsen, Simon F. ;
Coupe, Pierrick ;
Fonov, Vladimir ;
Manjon, Jose V. ;
Leung, Kelvin K. ;
Guizard, Nicolas ;
Wassef, Shafik N. ;
Ostergaard, Lasse Riis ;
Collins, D. Louis .
NEUROIMAGE, 2012, 59 (03) :2362-2373
[8]   Prostate segmentation by sparse representation based classification [J].
Gao, Yaozong ;
Liao, Shu ;
Shen, Dinggang .
MEDICAL PHYSICS, 2012, 39 (10) :6372-6387
[9]  
Gong D., 2010, 13 INT C ART INT STA
[10]  
Hahn HK, 2000, LECT NOTES COMPUT SC, V1935, P134