自閉症譜系障礙的早期檢測和治療有益於改善患者及其家人的生活品質。最近一項調查研究了過去十年中從磁振造影檢測各種腦部疾病的深度學習技術，使用自閉症大腦成像數據交換資料集中的靜息態功能性磁振造影的相關研究在診斷自閉症譜系障礙的分類器能夠獲得65.56％至70％的分類準確度。

而本論文主要宗旨為建立一個基於卷積神經網路的模型對大腦磁振造影進行自閉症疾病的分類。我們使用自閉症大腦成像數據交換的開放資料庫進行深度學習的模型建立，並使用資料集所提供的結構影像，先利用FMRIB Software Library進行影像的校正到MNI(Montreal Neurological Institute)立體定位空間中，再將影像由3D圖像轉為2D圖像，並將每個位置的2D圖像進行卷積神經網路的模型訓練，最後使用集成學習來提高對自閉症的辨識率。

最終我們的方法讓影像分類結果達到67.14％的分類準確率和73％的AUC ROC分數，這與過去使用靜息態功能性磁振造影的研究成果相去不遠，因此考慮到影像蒐集和前處理校正的複雜程度，若能使用結構影像進行分析，對於未來資料的收集也會較便利，在深度學習的研究也可以擁有用更多的影像來進行學習，來加強模型的一般性，對於自閉症研究有所幫助。

Early detection and treatments for Autism Spectrum Disorder (ASD) are beneficial in improving the life quality of a patient and his/her family. A recent survey studied deep learning techniques which have been developed over the past decade in detecting various brain diseases from MRIs. Some works which used rs-fMRI for diagnosing ASD were able to obtain diagnostics with an accuracy between 65.56% and 70%. In this research, instead of fMRI, a CNN-based model was proposed to identify reliable ASD-related biomarkers from sMRI. The sMRI images were spatially registered into the MNI stereotaxic space using FMRIB Software Library. In the work a 67.14% diagnosis accuracy and 73% AUC ROC were achieved, which were comparable to other studies using fMRI.

[1] L. Wan, M. Zeiler, S. Zhang, Y. Le Cun, and R. Fergus, “Regularization of neural networks using dropconnect,” in International Conference on Machine Learning, pp. 1058–1066, 2013.
[2] J. T. Springenberg, A. Dosovitskiy, T. Brox, and M. Riedmiller, “Striving for sim- plicity: The all convolutional net,” arXiv:1412.6806, 2014.
[3] J. Baio, “Prevalence of autism spectrum disorders: Autism and developmental dis- abilities monitoring network, 14 sites, united states, 2008. morbidity and mortality weekly report. surveillance summaries. volume 61, number 3,” Centers for Disease Control and Prevention, 2012.
[4] J.Baio,“Prevalenceofautismspectrumdisorderamongchildrenaged8years-autism and developmental disabilities monitoring network, 11 sites, united states, 2010,” 2014.
[5] D. L. Christensen, K. V. N. Braun, J. Baio, D. Bilder, J. Charles, J. N. Constantino, J. Daniels, M. S. Durkin, R. T. Fitzgerald, M. Kurzius-Spencer, et al., “Preva- lence and characteristics of autism spectrum disorder among children aged 8 years —autism and developmental disabilities monitoring network, 11 sites, united states, 2012,” MMWR Surveillance Summaries, vol. 65, no. 13, p. 1, 2018.
[6] J. Baio, L. Wiggins, D. L. Christensen, M. J. Maenner, J. Daniels, Z. Warren, M. Kurzius-Spencer, W. Zahorodny, C. R. Rosenberg, T. White, et al., “Prevalence of autism spectrum disorder among children aged 8 years—autism and developmen- tal disabilities monitoring network, 11 sites, united states, 2014,” MMWR Surveil- lance Summaries, vol. 67, no. 6, p. 1, 2018.
[7] M. D. Kogan, S. J. Blumberg, L. A. Schieve, C. A. Boyle, J. M. Perrin, R. M. Ghan- dour, G. K. Singh, B. B. Strickland, E. Trevathan, and P. C. van Dyck, “Prevalence of parent-reported diagnosis of autism spectrum disorder among children in the us, 2007,” Pediatrics, vol. 124, no. 5, pp. 1395–1403, 2009.
[8] J. Liu, M. Li, Y. Pan, W. Lan, R. Zheng, F.-X. Wu, and J. Wang, “Complex brain network analysis and its applications to brain disorders: a survey,” Complexity, vol. 2017, 2017.
[9] P. J. Basser, J. Mattiello, and D. LeBihan, “Estimation of the effective self-diffusion tensor from the nmr spin echo,” Journal of Magnetic Resonance, Series B, vol. 103, no. 3, pp. 247–254, 1994.
[10] S. Ogawa, T.-M. Lee, A. R. Kay, and D. W. Tank, “Brain magnetic resonance imag- ing with contrast dependent on blood oxygenation,” Proceedings of the National Academy of Sciences, vol. 87, no. 24, pp. 9868–9872, 1990.
[11] “Mri, fmri, and dti brain scan.” https://www.deviantart.com/ god-of-9-orphans/art/MRI-fMRI-DTI-747563567.
[12] R.Chen,Y.Jiao,andE.H.Herskovits,“Structuralmriinautismspectrumdisorder,” Pediatr Res, vol. 69, no. 5 Pt 2, pp. 63R–8R, 2011.
[13] A. L. Alexander, J. E. Lee, M. Lazar, R. Boudos, M. B. DuBray, T. R. Oakes, J. N. Miller, J. Lu, E.-K. Jeong, W. M. McMahon, et al., “Diffusion tensor imaging of the corpus callosum in autism,” Neuroimage, vol. 34, no. 1, pp. 61–73, 2007.
[14] B.G.Travers,D.P.Tromp,N.Adluru,N.Lange,D.Destiche,C.Ennis,J.A.Nielsen, A. L. Froehlich, M. B. Prigge, P. T. Fletcher, et al., “Atypical development of white matter microstructure of the corpus callosum in males with autism: a longitudinal investigation,” Molecular Autism, vol. 6, no. 1, p. 15, 2015.
[15] J. Radua, E. Via, M. Catani, and D. Mataix-Cols, “Voxel-based meta-analysis of regional white-matter volume differences in autism spectrum disorder versus healthy controls,” Psychological Medicine, vol. 41, no. 7, pp. 1539–1550, 2011.
[16] A. Tseng, Z. Wang, Y. Huo, S. Goh, J. A. Russell, and B. S. Peterson, “Differences in neural activity when processing emotional arousal and valence in autism spectrum disorders,” Human Brain Mapping, vol. 37, no. 2, pp. 443–461, 2016.
[17] B. Yegnanarayana, Artificial Neural Networks. PHI Learning Pvt. Ltd., 2009.
[18] J. Schmidhuber, “Deep learning in neural networks: An overview,” Neural Net- works, vol. 61, pp. 85–117, 2015.
[19] H.BourlardandY.Kamp,“Auto-associationbymultilayerperceptronsandsingular value decomposition,” Biological Cybernetics, vol. 59, no. 4-5, pp. 291–294, 1988.
[20] G. Alain and Y. Bengio, “What regularized auto-encoders learn from the data- generating distribution,” The Journal of Machine Learning Research, vol. 15, no. 1, pp. 3563–3593, 2014.
[21] Y. LeCun, Y. Bengio, et al., “Convolutional networks for images, speech, and time series,” The Handbook of Brain Theory and Neural Networks, vol. 3361, no. 10, p. 1995, 1995.
[22] Y. LeCun, K. Kavukcuoglu, and C. Farabet, “Convolutional networks and applica- tions in vision,” in Proceedings of 2010 IEEE International Symposium on Circuits and Systems, pp. 253–256, IEEE, 2010.
[23] Y. LeCun, L. Bottou, Y. Bengio, P. Haffner, et al., “Gradient-based learning applied to document recognition,” Proceedings of the IEEE, vol. 86, no. 11, pp. 2278–2324, 1998.
[24] P.Y.Simard,D.Steinkraus,J.C.Platt,etal.,“Bestpracticesforconvolutionalneural networks applied to visual document analysis,” in Icdar, vol. 3, 2003.
[25] K. Simonyan and A. Zisserman, “Very deep convolutional networks for large-scale image recognition,” arXiv:1409.1556, 2014.
[26] K.He,X.Zhang,S.Ren,andJ.Sun,“Deepresiduallearningforimagerecognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 770–778, 2016.
[27] K.He,X.Zhang,S.Ren,andJ.Sun,“Identitymappingsindeepresidualnetworks,” in European conference on computer vision, pp. 630–645, Springer, 2016.
[28] T. Greevink, “Prioritized experience replay based on the wasserstein metric in deep reinforcement learning: The regularizing effect of modelling return distributions,” 2019.
[29] H.-C. Shin, M. R. Orton, D. J. Collins, S. J. Doran, and M. O. Leach, “Stacked autoencoders for unsupervised feature learning and multiple organ detection in a pilot study using 4d patient data,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 35, no. 8, pp. 1930–1943, 2012.
[30] M. Ghafoorian, N. Karssemeijer, T. Heskes, M. Bergkamp, J. Wissink, J. Obels, K. Keizer, F.-E. de Leeuw, B. van Ginneken, E. Marchiori, et al., “Deep multi-scale location-aware 3d convolutional neural networks for automated detection of lacunes of presumed vascular origin,” NeuroImage: Clinical, vol. 14, pp. 391–399, 2017.
[31] Q.Dou,H.Chen,L.Yu,L.Zhao,J.Qin,D.Wang,V.C.Mok,L.Shi,andP.-A.Heng, “Automatic detection of cerebral microbleeds from mr images via 3d convolutional neural networks,” IEEE Transactions on Medical Imaging, vol. 35, no. 5, pp. 1182– 1195, 2016.
[32] G.Wu,M.Kim,Q.Wang,Y.Gao,S.Liao,andD.Shen,“Unsuperviseddeepfeature learning for deformable registration of mr brain images,” in International Conference on Medical Image Computing and Computer Assisted Intervention, pp. 649–656, Springer, 2013.
[33] G. Wu, M. Kim, Q. Wang, B. C. Munsell, and D. Shen, “Scalable high-performance image registration framework by unsupervised deep feature representations learn- ing,” IEEE Transactions on Biomedical Engineering, vol. 63, no. 7, pp. 1505–1516, 2015.
[34] “Ischemicstrokelesionsegmentation.”http//www.isles-challenge.org/.
[35] “Multimodal brain tumor segmentation challenge 2019.” http://
braintumorsegmentation.org/.
[36] “Grand challenge on mr brain segmentation at miccai 2018.” https://
mrbrains18.isi.uu.nl/.
[37] “Left ventricular segmentation challenge.” http://www.cardiacatlas.org/
challenges/lv-segmentation-challenge/.
[38] W. Zhang, R. Li, H. Deng, L. Wang, W. Lin, S. Ji, and D. Shen, “Deep convolu- tional neural networks for multi-modality isointense infant brain image segmenta- tion,” NeuroImage, vol. 108, pp. 214–224, 2015.
[39] P. Moeskops, M. A. Viergever, A. M. Mendrik, L. S. de Vries, M. J. Benders, and I. Išgum, “Automatic segmentation of mr brain images with a convolutional neural network,” IEEE Transactions on Medical Imaging, vol. 35, no. 5, pp. 1252–1261, 2016.
[40] S.Pereira,A.Pinto,V.Alves,andC.A.Silva,“Braintumorsegmentationusingcon- volutional neural networks in mri images,” IEEE Transactions on Medical Imaging, vol. 35, no. 5, pp. 1240–1251, 2016.
[41] M. Havaei, A. Davy, D. Warde-Farley, A. Biard, A. Courville, Y. Bengio, C. Pal, P.-M. Jodoin, and H. Larochelle, “Brain tumor segmentation with deep neural net- works,” Medical Image analysis, vol. 35, pp. 18–31, 2017.
[42] S. Ghiassian, R. Greiner, P. Jin, and M. R. Brown, “Using functional or structural magnetic resonance images and personal characteristic data to identify adhd and autism,” PloS one, vol. 11, no. 12, p. e0166934, 2016.
[43] H.-I.SukandD.Shen,“Deeplearning-basedfeaturerepresentationforad/mciclassi- fication,” in International Conference on Medical Image Computing and Computer Assisted Intervention, pp. 583–590, Springer, 2013.
[44] S. Sarraf, G. Tofighi, et al., “Deepad: Alzheimer′ s disease classification via deep convolutional neural networks using mri and fmri,” BioRxiv, p. 070441, 2016.
[45] E. Hosseini-Asl, R. Keynton, and A. El-Baz, “Alzheimer’s disease diagnostics by adaptation of 3d convolutional network,” in 2016 IEEE International Conference on Image Processing (ICIP), pp. 126–130, IEEE, 2016.
[46] H.-I. Suk, S.-W. Lee, and D. Shen, “Hierarchical feature representation and mul- timodal fusion with deep learning for ad/mci diagnosis,” NeuroImage, vol. 101, pp. 569–582, 2014.
[47] H.-I. Suk, S.-W. Lee, D. Shen, A. D. N. Initiative, et al., “Latent feature representa- tion with stacked auto-encoder for ad/mci diagnosis,” Brain Structure and Function, vol. 220, no. 2, pp. 841–859, 2015.
[48] S. Liu, S. Liu, W. Cai, S. Pujol, R. Kikinis, and D. Feng, “Early diagnosis of alzheimer’s disease with deep learning,” in 2014 IEEE 11th International Sympo- sium on Biomedical Imaging (ISBI), pp. 1015–1018, IEEE, 2014.
[49] A. Di Martino, C.-G. Yan, Q. Li, E. Denio, F. X. Castellanos, K. Alaerts, J. S. An- derson, M. Assaf, S. Y. Bookheimer, and M. Dapretto, “The autism brain imaging data exchange: Towards a large-scale evaluation of the intrinsic brain architecture in autism,” Molecular psychiatry, vol. 19, no. 6, p. 659, 2014.
[50] “Autism brain imaging data exchange i.” http://fcon_1000.projects.nitrc. org/indi/abide/manuscripts.html.
[51] “Fmrib software library v6.0.” https://fsl.fmrib.ox.ac.uk/fsl/fslwiki.
[52] M. Jenkinson and S. Smith, “A global optimisation method for robust affine regis-
tration of brain images,” Medical Image Analysis, vol. 5, no. 2, pp. 143–156, 2001.
[53] M. Jenkinson, P. Bannister, M. Brady, and S. Smith, “Improved optimization for the robust and accurate linear registration and motion correction of brain images,” Neuroimage, vol. 17, no. 2, pp. 825–841, 2002.
[54] V. Fonov, A. C. Evans, K. Botteron, C. R. Almli, R. C. McKinstry, D. L. Collins, B. D. C. Group, et al., “Unbiased average age-appropriate atlases for pediatric stud- ies,” Neuroimage, vol. 54, no. 1, pp. 313–327, 2011.
[55] V. S. Fonov, A. C. Evans, R. C. McKinstry, C. Almli, and D. Collins, “Unbiased nonlinear average age-appropriate brain templates from birth to adulthood,” Neu- roImage, no. 47, p. S102, 2009.
[56] F. Rosenblatt, “The perceptron: a probabilistic model for information storage and organization in the brain,” Psychological review, vol. 65, no. 6, p. 386, 1958.
[57] D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv: 1412.6980, 2014.
[58] S. Ioffe and C. Szegedy, “Batch normalization: Accelerating deep network training by reducing internal covariate shift,” arXiv:1502.03167, 2015.
[59] B. Zhou, A. Khosla, A. Lapedriza, A. Oliva, and A. Torralba, “Learning deep fea- tures for discriminative localization,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 2921–2929, 2016.
[60] R.R.Selvaraju,M.Cogswell,A.Das,R.Vedantam,D.Parikh,andD.Batra,“Grad- cam: Visual explanations from deep networks via gradient-based localization,” in Proceedings of the IEEE International Conference on Computer Vision, pp. 618– 626, 2017.
[61] S. Clare, “Functional mri: Methods and applications,” University of Nottingham, p. 155, 1997.
[62] A. Jaiswal, “Large-scale unsupervised deep representation learning for brain struc- ture,” arXiv:1805.01049, 2018.
[63] M. A. Aghdam, “Combination of rs-fmri and smri data to discriminate autism spec- trum disorders in young children using deep belief network,” Journal of Digital Imaging, vol. 31, no. 6, pp. 895–903, 2018.
[64] A.S.Heinsfeld,“Identificationofautismspectrumdisorderusingdeeplearningand the abide dataset,” NeuroImage: Clinical, vol. 17, pp. 16–23, 2018.
[65] X. Li, N. C. Dvornek, J. Zhuang, P. Ventola, and J. S. Duncan, “Brain biomarker interpretation in asd using deep learning and fmri,” in International Conference on Medical Image Computing and Computer Assisted Intervention, pp. 206–214, Springer, 2018.
[66] N. C. Dvornek, “Combining phenotypic and resting-state fmri data for autism clas- sification with recurrent neural networks,” in 2018 IEEE 15th International Sympo- sium on Biomedical Imaging (ISBI 2018), pp. 725–728, IEEE, 2018.
[67] S.Parisot,S.I.Ktena,E.Ferrante,M.Lee,R.G.Moreno,B.Glocker,andD.Rueck- ert, “Spectral graph convolutions for population-based disease prediction,” in International Conference on Medical Image Computing and Computer Assisted Interven- tion, pp. 177–185, Springer, 2017.
[68] M. Ismail, G. Barnes, M. Nitzken, A. Switala, A. Shalaby, E. Hosseini-Asl, M. Casanova, R. Keynton, A. Khalil, and A. El-Baz, “A new deep-learning approach for early detection of shape variations in autism using structural mri,” in 2017 IEEE International Conference on Image Processing (ICIP), pp. 1057–1061, IEEE, 2017.