Application of Machine Learning in Diagnosis and Prediction of Neuropsychiatric Diseases
-
摘要: 多模态磁共振神经影像技术的发展为大脑工作原理研究及脑部疾病早期诊断提供了新的手段。当前多数神经精神性疾病的诊断仅依据其临床症状, 缺少客观的神经影像生物学标志物。传统的基于组间比较的单变量分析仅能在组间水平进行推断, 无法提供个体水平的诊断和预测, 临床应用价值非常有限。机器学习技术可在不同水平对神经影像进行计算分析和研究, 发现其中规律从而有效预测和分类未知数据, 找出与脑疾病高度相关的脑区特征, 提供个体水平的诊断并探测脑疾病的病理生理机制。本文对基于机器学习的神经影像分析步骤及机器学习在神经精神疾病智能诊断及预测研究中的应用进行综述, 并对未来研究发展进行展望。Abstract: The development of multi-modal magnetic resonance imaging (MRI) provides a new method for early diagnosis of brain diseases. Currently, diagnosis of most neuropsychiatric diseases is only based on clinical symptoms, which lacks objective neuroimaging biomarkers. As univariate analysis that is used in traditional research can only reveal disease-related structural and functional alterations at group level, which limits the clinical application. Recent attention has turned toward integrating multi-modal neuroimaging and machine learning to assist clinical disease diagnosis. Machine learning can obtain rules via automatically analyzing neuro-imaging data, and apply these rules to predict unknown data, to identify the brain regions highly correlated to the brain disease, to provide individual levels of diagnosis, and to detect pathophysiological mechanisms. This paper reviews the concrete steps of neuroimaging analysis based on machine learning and the application of machine learning in intelligent diagnosis and prediction of neuropsychiatric diseases. Finally, some new directions for future research are forecasted.
-
[1] He Y, Chen Z, Evans A. Structural insights into aberrant topological patterns of large-scale cortical networks in Alzheimer's disease[J]. J Neurosci, 2008, 4:4756-4766. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=1d84f46d0ea2a3673944853360ffb987 [2] Lo CY, Wang PN, Chou KH, et al. Diffusion tensor tractography reveals abnormal topological organization in structural cortical networks in Alzheimer's disease[J]. J Neurosci, 2010, 30:16876-16885. doi: 10.1523/JNEUROSCI.4136-10.2010 [3] Cao Q, Shu N, An L, et al. Probabilistic diffusion tractography and graph theory analysis reveal abnormal white matter structural connectivity networks in drug-naive boys with attention deficit/hyperactivity disorder[J]. J Neurosci, 2013, 33:10676-10687. doi: 10.1523/JNEUROSCI.4793-12.2013 [4] Wen H, Liu Y, Wang J, et al. Combining tract- and atlas-based analysis reveals microstructural abnormalities in early Tourette syndrome children[J]. Hum Brain Mapp, 2016, 37:1903-1919. doi: 10.1002/hbm.23146 [5] Dai D, Wang J, Hua J, et al. Classification of ADHD children through multimodal magnetic resonance imaging[J]. Front Syst Neurosci, 2012, 6:63-63. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=Doaj000003592374 [6] Wen H, Liu Y, Rekik I, et al. Multi-modal multiple kernel learning for accurate identification of Tourette syndrome children[J]. Pattern Recognit, 2017, 63:601-611. doi: 10.1016/j.patcog.2016.09.039 [7] Wen H, Liu Y, Rekik I, et al. Disrupted topological organization of structural networks revealed by probabilistic diffusion tractography in Tourette syndrome children[J]. Hum Brain Mapp, 2017, 38:3988-4008. doi: 10.1002/hbm.23643 [8] Zang Y, Jiang T, Lu Y, et al. Regional homogeneity approach to fMRI data analysis[J]. Neuroimage, 2004, 22:394-400. doi: 10.1016/j.neuroimage.2003.12.030 [9] Zang YF, He Y, Zhu CZ, et al. Altered baseline brain activity in children with ADHD revealed by resting-state functional MRI[J]. Brain Dev, 2007, 29:83-91. doi: 10.1016/j.braindev.2006.07.002 [10] Wen H, Liu Y, Rekik I, et al. Combining disrupted and discriminative topological properties of functional connectivity networks as neuroimaging biomarkers for accurate diagnosis of early tourette syndrome children[J]. Mol Neurobiol, 2017. doi: 10.1007/s12035-017-0519-1.[Epub ahead of print]. [11] Breiman L. Bagging predictors[J]. Mach Learn, 1996, 24:123-140. doi: 10.1023/A:1018054314350 [12] Schapire RE, Freund Y, Bartlett P, et al. Boosting the margin:a new explanation for the effectiveness of voting methods[J]. Ann Stat, 1998, 26:1651-1686. doi: 10.1214/aos/1024691352 [13] Wilson SM, Ogar JM, Laluz V, et al. Automated MRI-based classification of primary progressive aphasia variants[J]. Neuroimage, 2009, 47:1558-1567. doi: 10.1016/j.neuroimage.2009.05.085 [14] Hinton GE, Salakhutdinov RR. Reducing the dimensionality of data with neural networks[J]. Science, 2006, 313:504. doi: 10.1126/science.1127647 [15] 孙志军, 薛磊, 许阳明, 等.深度学习研究综述[J].计算机应用研究, 2012, 29:2806-2810. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=jsjyyyj201208002 [16] Zou X, Wang G, Yu G. Protein function prediction using deep restricted boltzmann machines[J]. Biomed Res Int, 2017, 2017:1729301. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=b38f7320bfe73c1058e13eba56309f64 [17] Hinton GE. Deep belief networks[J]. Scholarpedia, 2009, 4:5947. doi: 10.4249/scholarpedia.5947 [18] Graves A, Mohamed A, Hinton GE. Speech recognition with deep recurrent neural networks[C]. IEEE International Conference on Acoustics, Speech and Signal Processing, 2013: 6645-6649. [19] Krizhevsky A, Sutskever I, Hinton GE. ImageNet classific-ation with deep convolutional neural networks[C]. Advances in Neural Information Processing Systems, 2012: 1097-1105. [20] Esteva A, Kuprel B, Novoa RA, et al. Dermatologist-level classification of skin cancer with deep neural networks[J]. Nature, 2017, 542:115. doi: 10.1038/nature21056 [21] Gulshan V, Peng L, Coram M, et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs[J]. JAMA, 2016, 316:2402. doi: 10.1001/jama.2016.17216 [22] Dai D, He H, Vogelstein JT, et al. Accurate prediction of AD patients using cortical thickness networks[J]. Mach Vis Appl, 2013, 24:1445-1457. doi: 10.1007/s00138-012-0462-0 [23] 张超, 胡斌, 郑炜豪.脑皮层形态学多特征融合的轻度认知障碍的分类研究[J].小型微型计算机系统, 2016, 37:2558-2561. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=xxwxjsjxt201611033 [24] 李慧卓, 相洁, 秦嘉玮, 等.基于Adaboost的轻度认知障碍和阿尔茨海默病分类[J].中国医学影像技术, 2016, 32:623-627. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=zgyxyxjs201604037 [25] 吕鸿蒙, 赵地, 迟学斌.基于增强AlexNet的深度学习的阿尔茨海默病的早期诊断[J].计算机科学, 2017, (S1):50-60. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=jsjkx2017z1011 [26] Mueller SG, Weiner MW, Thal LJ, et al. The Alzheimer's disease neuroimaging initiative[J]. Neuroimaging Clin N Am, 2005, 15:869. doi: 10.1016/j.nic.2005.09.008 [27] Consortium H. The ADHD-200 Consortium:A model to advance the translational potential of neuroimaging in clinical neuroscience[J]. Front Syst Neurosci, 2012, 6:62. https://www.ncbi.nlm.nih.gov/pubmed/22973200 [28] 朱莉, 张丽英, 韩云涛, 等.基于卷积神经网络的注意缺陷多动障碍分类研究[J].生物医学工程学杂志, 2017, (1):99-105. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=swyxgcx201701018 [29] Kuang D, Guo X, An X, et al. Discrimination of ADHD based on fMRI data with deep belief network[C]. International Conference on Intelligent Computing, 2014: 225-232. [30] Greene DJ, Church JA, Dosenbach NUF, et al. Multivariate pattern classification of pediatric tourette syndrome using functional connectivity MRI[J]. Dev Sci, 2016, 19:581-598. doi: 10.1111/desc.12407 [31] Liao W, Yu Y, Miao HH, et al. Inter-hemispheric intrinsic connectivity as a neuromarker for the diagnosis of boys with tourette syndrome[J]. Mol Neurobiol, 2017, 54:2781-2789. doi: 10.1007/s12035-016-9863-9 [32] Wen H, Liu Y, Wang S, et al. Multi-threshold white matter structural networks fusion for accurate diagnosis of early tou-rette syndrome children[C]. International Society for Optics and Photonics, 2017: 101341Q-101341Q-13. [33] Wang B, Mezlini AM, Demir F, et al. Similarity network fusion for aggregating data types on a genomic scale[J]. Nat Methods, 2014, 11:333-337. doi: 10.1038/nmeth.2810 [34] Qi J, Tejedor J. Deep multi-view representation learning for multi-modal features of the schizophrenia and Schizo-affective disorder[C]. IEEE International Conference on Acoustics, Speech and Signal Processing, 2016: 116-118. [35] Adeli E, Wu G, Saghafi B, et al. Kernel-based joint feature selection and Max-Margin classification for early diagnosis of Parkinson's disease[J]. Sci Rep, 2017, 7:41069. doi: 10.1038/srep41069 [36] Crippa A, Salvatore C, Perego P, et al. Use of machine learning to identify children with autism and their motor abnormalities[J]. J Autism Dev Disord, 2015, 45:2146-2156. doi: 10.1007/s10803-015-2379-8 [37] Wang J, Wang Q, Peng J, et al. Multi-task diagnosis for autism spectrum disorders using multi-modality features:a multi-center study[J]. Hum Brain Mapp, 2017, 38:3081-3097. doi: 10.1002/hbm.23575 [38] Chekroud AM, Zotti RJ, Shehzad Z, et al. Cross-trial prediction of treatment outcome in depression:a machine learning approach[J]. Lancet Psychiatry, 2016, 3:243. doi: 10.1016/S2215-0366(15)00471-X [39] Hoexter MQ, Miguel EC, Diniz JB, et al. Predicting obsessive-compulsive disorder severity combining neuroimaging and machine learning methods[J]. J Affect Disord, 2013, 150:1213-1216. doi: 10.1016/j.jad.2013.05.041 [40] 陈诗慧, 刘维湘, 秦璟, 等.基于深度学习和医学图像的癌症计算机辅助诊断研究进展[J].生物医学工程学杂志, 2017, (2):314-319. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=swyxgcx201702024 [41] Huys QJM, Maia TV, Frank MJ. Computational psychiatry as a bridge from neuroscience to clinical applications[J]. Nat Neurosci, 2016, 19:404-413. doi: 10.1038/nn.4238
点击查看大图
计量
- 文章访问数: 348
- HTML全文浏览量: 63
- PDF下载量: 810
- 被引次数: 0