Automated lesion detection on MRI scans using combined unsupervised and supervised methods

BMC Medical Imaging

Song Wang²  Kang Zheng²  Hongkai Yu²  Christopher Rorden¹  Paul Fillmore¹  Julius Fridriksson¹  Dazhou Guo² 

[1]Department of Communication Science & Disorders, University of South Carolina, 915 Greene Street, Columbia 29208, USA

[2]Department of Computer Science & Engineering, University of South Carolina, 301 Main Street, Columbia 29201, USA

关键词: Unsupervised and supervised methods;    Magnetic resonance imaging (MRI);    Lesion detection;   

Others  :  1233374
DOI  :  10.1186/s12880-015-0092-x

 received in 2015-05-12, accepted in 2015-10-16,  发布年份 2015

【 摘 要 】

Background

Accurate and precise detection of brain lesions on MR images (MRI) is paramount for accurately relating lesion location to impaired behavior. In this paper, we present a novel method to automatically detect brain lesions from a T1-weighted 3D MRI. The proposed method combines the advantages of both unsupervised and supervised methods.

Methods

First, unsupervised methods perform a unified segmentation normalization to warp images from the native space into a standard space and to generate probability maps for different tissue types, e.g., gray matter, white matter and fluid. This allows us to construct an initial lesion probability map by comparing the normalized MRI to healthy control subjects. Then, we perform non-rigid and reversible atlas-based registration to refine the probability maps of gray matter, white matter, external CSF, ventricle, and lesions. These probability maps are combined with the normalized MRI to construct three types of features, with which we use supervised methods to train three support vector machine (SVM) classifiers for a combined classifier. Finally, the combined classifier is used to accomplish lesion detection.

Results

We tested this method using T1-weighted MRIs from 60 in-house stroke patients. Using leave-one-out cross validation, the proposed method can achieve an average Dice coefficient of 73.1 % when compared to lesion maps hand-delineated by trained neurologists. Furthermore, we tested the proposed method on the T1-weighted MRIs in the MICCAI BRATS 2012 dataset. The proposed method can achieve an average Dice coefficient of 66.5 % in comparison to the expert annotated tumor maps provided in MICCAI BRATS 2012 dataset. In addition, on these two test datasets, the proposed method shows competitive performance to three state-of-the-art methods, including Stamatakis et al., Seghier et al., and Sanjuan et al.

Conclusions

In this paper, we introduced a novel automated procedure for lesion detection from T1-weighted MRIs by combining both an unsupervised and a supervised component. In the unsupervised component, we proposed a method to identify lesioned hemisphere to help normalize the patient MRI with lesions and initialize/refine a lesion probability map. In the supervised component, we extracted three different-order statistical features from both the tissue/lesion probability maps obtained from the unsupervised component and the original MRI intensity. Three support vector machine classifiers are then trained for the three features respectively and combined for final voxel-based lesion classification.

【 授权许可】

   
2015 Guo et al.

【 图 表 】

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Fig. 7.

Fig. 8.

Fig. 9.

Fig. 10.

【 参考文献 】

• [1]Fridriksson J, Guo D, Fillmore P, Holland A, Rorden C. Damage to the anterior arcuate fasciculus predicts non-fluent speech production in aphasia. Brain. 2013; 136:3451-460.
• [2]Tecichmann M, Kas A, Boutet C, Ferrieux S, Nogues M, Samri D et al.. Deciphering logopenic primary progressive aphasia: A clinical, imaging and biomarker investigation. Brain. 2013; 136:3474-488.
• [3]Smith DV, Clithero JA, Rorden C, Karnath H. Decoding the anatomical network of spatial attention. PNAS. 2013; 110:1518-1523.
• [4]Fiebach J, Jansen O, Schellinger P, Knauth M, Hartmann M, Heiland S et al.. Comparison of ct with diffusion-weighted mri in patients with hyperacute stroke. Neuroradiol. 2001; 43:628-32.
• [5]Farr TD, Wegener S. Use of magnetic resonance imaging to predict outcome after stroke: A review of experimental and clinical evidence. J Cereb Blood Flow Metab. 2010; 30:703-17.
• [6]Wilke M, de Haan B, Juenger H, Karmath HO. Manual, semi-automated, and automated delineation of chronic brain lesions: a comparison of methods. Neuroimage. 2011; 56:2038-046.
• [7]Anbeek P, Vincken KL, Van Osch MJ, Bisschops RH, Van der Ground J. Automatic segmentation of different-sized white matter lesions by voxel probability estimation. Med Image Anal. 2004; 8:205-15.
• [8]Bernasconi A. Advanced mri analysis methods for detection of focal cortical dysplasia. Epileptic Disord. 2003; 5:81-4.
• [9]Bilello M, Arkuszewski M, Nucifora P, Nasrallah I, Melhem ER, Krejza J. Multiple sclerosis: Identification of temporal changes in brain lesions with computer-assisted detection software. Neuroradiol J. 2013; 26:143-50.
• [10]Ong HK, Ramachandram D, Mandava R, Shuaib IL. Automatic white matter lesion segmentation using an adaptive outlier detection method. Neuroimage. 2012; 30:807-23.
• [11]Seghier ML, Ramlackhansingh A, Crinion J, Leff AP, Price CJ. Lesion identification using unified segmentation-normalization models and fuzzy clustering. Neuroimage. 2008; 41(4):1253-1266.
• [12]Spies L, Tewes A, Opfer R, Buchert R, Winkler G et al.. Fully automatic detection of deep white matter t1 hypointense lesions in multiple sclerosis. Phys Med Biol. 2013; 58:8323-337.
• [13]Stamatakis EA, Tyler LK. Identifying lesions on structural brain images-validation of the method and application to neuropsychological patients. Brain and Lang. 2005; 94:167-77.
• [14]Artzi M, Aizenstein O, Jonas-Kimchi T, Myers V, Hallevi H, Bashat DB. Flari lesion segmentation: Application in patients with brain tumors and acute ischemic stroke. Eur J Radiol. 2013; 82:1512-1518.
• [15]Despotovic I, Segers I, Platisa L, Vansteenkiste E, Plzurica A, Deblaere K, et al.Automatic 3d graph cuts for brain cortex segmentation in patients with focal cortical dysplasia. Conf Proc IEEE Eng Med Biol Soc. 2011;:7981–4.
• [16]Gondal AH, Khan MNA. A review of fully automated techniques for brain tumor detection from mr images. Int J Comput Sci. 2013; 2:55-61.
• [17]Gordillo N, Montseny E, Sobrevilla P. State of the art survey on mri brain tumor segmentation. Magn Reson Imaging. 2013; 31:1426-1438.
• [18]Shen S, Szameitat AJ, Sterr A. Detection of infarct lesions from single mri modality using inconsistency between voxel intensity and spatial location - a 3-d automatic approach. IEEE Trans Inf Technol Biomed. 2008; 12:532-40.
• [19]Shen S, Szameitat AJ, Sterr A. An improved lesion detection approach based on similarity measurement between fuzzy intensity segmentation and spatial probability maps. Magn Reson Imaging. 2010; 28:245-54.
• [20]Sanjun A, Price CJ, Mancini L, Josse G, Grogan A, Yamamoto AK, et al.Automated identification of brain tumors from single mr images based on segmentation with refined patient-specific priors. Front Neurosci. 2013.
• [21]Xu T, Mandal M. Automatic brain tumor extraction from t1-weighted coronal mri using fast bounding box and dynamic snake. International Conference of the IEEE Engineering in Medicine and Biology Society. 2012:444–447.
• [22]Abdulah BA, Younis AA, John NM. Multi-sectional views textural based svm for ms lesion segmentation in multi-channels mris. The Open Biomed Eng J. 2012; 6:56-72.
• [23]Desikan R, Cabral H, Hess C et al.. Automated MRI measures identify individuals with mild cognitive impairment and Alzheimers disease. Brain. 2009; 132:2048-057.
• [24]Fiot JB, Cohen LD, Raniga P, Fripp J. Efficient brain lesion segmentation using multi-modality tissue-based feature selection and support vector machines. Int J Numer Methods in Biomed Eng. 2013; 29:905-15.
• [25]Geremia E, Clatz O, Menze BH, Konukoglu E, Criminisi A, Ayache N. Spatial decision forests for ms lesion segmentation in multi-channel magnetic resonance images. NeuroImage. 2011; 57:378-90.
• [26]Lao Z, Shen D, Liu D, Jawad AF, Melhem ER, Launer LJ et al.. Computer-assisted segmentation of white matter lesions in 3d mr images using support vector machine. Acad Radiol. 2008; 15:300-13.
• [27]Morra J, Tu Z, Toga A, Thompson P. Automated segmentation of ms lesions using a contextual model for the miccai grand challenge. MICCAI 2008 Workshop. 2008.
• [28]Quddus A, Fieguth P, Basir O. Adaboost and support vector machines for white matter lesion segmentation in mr images. Conf IEEE Eng Med Biol Soc. 2005; 1:463-6.
• [29]Schneell S, Saur D, Kreher BW, Hennig J, Burkhardt H, Kiselev VG. Fully automated classification of hardi in vivo data using a support vector machine. NeuroImage. 2009; 46:642-51.
• [30]Ashburner J, Friston KJ. Unified segmentation. NeuroImage. 2005; 26:839-51.
• [31]Seghier ML, Friston KJ, Price CJ. Detecting subject-specific activation using fuzzy clustering. Neuroimage. 2007; 36(3):594-605.
• [32]Breiman L. Random forests. Machine Learning. 2001; 45:5-32.
• [33]Kovesi P. Image features from phase congruency. Videre: J Comp Vis Res. 1999; 1:1-26.
• [34]Jiang J, Wu Y, Huang M, Yang W, Chen W, Feng Q. 3d brain tumor segmentation in multimodal mr images based on learning population and patient specific feature sets. Comput Med Imaging Graph. 2013; 37:512-21.
• [35]Popuri K, Cobzas D, Murtha A, Jagersand M. 3d variational brain tumor segmentation using a high dimensional feature set. In J Comput Assisted Radiology. 2012; 7:493-506.
• [36]Alipanahi B, Biggs M, Ghodsi A. Distance metric learning vs. fisher discriminant analysis. In: Cohn A, editor. Proceedings of the 23rd national conference on Artificial intelligence - Volume 2 (AAAI’08), Vol. 2. AAAI Press;. 2008. p. 598-603.
• [37]Aprile I, Iaiza F, Lavaroni A, Budai R, Dolso P, Scott CA et al.. Analysis of cystic intracranial lesions performed with fluid-attenuated inversion recovery mr imaging. Am J Neuroradiology. 1999; 20:1259-1267.
• [38]Ohe Y, Hayashi T, Deguchi I, Fukuoka T, Horiuchi Y, Maruyama H et al.. Mri abnormality of the pulvinar in patients with status epilepticus. J Neuroradiol. 2014; 41:220-6.
• [39]Nachev P, Coulthard E, Jager HR, Kennard C, Husain M. Enantiomorphic normalization of focally lesioned brains. Neuroimage. 2008; 39(3):1215-1226.
• [40]Evans AC, Collins DL, Millst SR, Brown ED, Kelly RL, Peters TM. 3d statistical neuroanatomical models from 305 mri volumes. Nucl Sci Symp Med Imaging Conf. 1993; 3:1813-1817.
• [41]SPM: Statistical Parametric Mapping. http://www.fil.ion.ucl.ac.uk/spm/. Accessed Apr 2009.
• [42]Breet M, Leff AP, Rorden C, Ashburner J. Spatial normalization of brain images with focal lesions using cost function masking. Neuroimage. 2001; 14:486-500.
• [43]Toga AW, Thompson PM. Mapping brain asymmetry. Nat Rev Neurosci. 2003; 4:37-48.
• [44]SPM, 8 Scripts. http://www. mccauslandcenter.sc.edu/CRNL/tools/spm8-scripts webcite
• [45]Collignon A, Maes F, Delaere D, Vandermeulen D, Suetens P, Marchal G. Automated multi-modality image registration based on information theory. Inf Process Med Imaging. 1995; 3:263-74.
• [46]Tyler LK, Marslen-Wilson W, Stamatakis EA. Dissociating neuro-cognitive component processes: voxel-based correlational methodlogy. Neuropsychologia. 2005;:771–8.
• [47]Larsen R. An Introduction to Mathematical Statistics and Its Applications. Prentice Hall, Upper Saddle River, N; 2001.
• [48]Friston KJ, Josephs O, Zarahn E, Holmes AP, Rouquette S, Poline JB. To smooth or not to smooth. NeuroImage. 2000; 12:196-208.
• [49]Cui X, Liu Y, Shan S, Chen X, Gao W. 3d haar-like features for pedestrian detection. 2007 IEEE Int Conf Multimed Expo. 2007; 1:1263-1266.
• [50]Cortes C, Vapnik V. Support-vector networks. Mach Learn. 1995; 20:273-97.
• [51]Lu J, Wang G, Moulin P. Image set classification using holistic multiple order statistics features and localized multi-kernel metric learning. In Proceedings of the 2013 IEEE International Conference on Computer Vision (ICCV ’13). Washington, DC, USA: IEEE Computer Society: 2013. p. 329–336, doi:. http://dx. [10.1109/ICCV.2013.48] webcitedoi.org/10.1109/ICCV.2013.48 webcite
• [52]Rorden C, Brett M. Stereotaxic display of brain lesions. Behav Neurol. 2000; 12:191-200.
• [53]Lee MA, Smith S, Palace J, Mathews PM. Defining multiple sclerosis disease activity using mri t2-weighted difference imaging. Brain. 1998; 121:2095-102.
• [54]Jovicich J, Czanner S, Greve D, Haley E, Kouwe A, Gollub R et al.. Reliability in multi-site structural mri studies: Effects of gradient non-linearity correction on phantom and human data. NeuroImage. 2006; 30:436-43.
• [55]Liu Y, Collins R. Robust midsagittal plane extraction from normal and pathological 3d neuroradiology images. IEEE Transact Med Imaging. 2001; 20:173-92.
• [56]Tuzikov AV, Colliot O, Block I. Evaluation of the symmetry plane in 3d mr images. Pattern Recog Lettters. 2003; 24:2219-233.