-
阿尔兹海默症(Alzheimer disease,AD)是一种老年性神经退化性疾病,主要临床表现为认知功能下降、记忆力逐渐丧失、精神和日常生活能力显著降低。目前全球AD患者超过千万,且预计在未来的几十年内其患病率将急剧增长[1]。
AD作为一种不可逆神经退行性疾病,发病机制十分复杂,涉及到病理、生理系统间的各种相互作用,主要的病因包括脑灰质中β-淀粉样蛋白积聚的淀粉样蛋白斑块和与tau蛋白相关的神经纤维缠结[2]。有研究表明,AD的疾病进程与其他类型的分子病理、生理机制也存在相关性,如突触功能障碍、神经炎症和脑代谢功能紊乱等[3-5]。因此,明确AD的发病进程和病理、生理学特征是一项十分艰巨的挑战。
近年来随着人工智能技术的兴起,各种数据挖掘技术被广泛地应用于AD早期诊断和对其机理的研究中,其中,基于图论的复杂网络分析方法为临床工作者提供了一个全新的视角[6]。该方法将大脑看作一个复杂的交互网络系统,将脑区或神经元组视作网络节点,通过对比观察AD患者和正常人群中脑网络参数变化,实现对AD的早期鉴别和机理研究。在过去的十几年,从动物模型到活体人脑的跨物种研究已经揭示了AD会造成脑网络结构和功能变化,如小世界特性、富人俱乐部和分层模块化等特性的变化[7-10]。而随着诸多神经影像学技术[如结构磁共振成像(structural magnetic resonance imaging,sMRI)、功能磁共振成像(functional magnetic resonance imaging, fMRI)、PET等]在AD临床的大规模普及应用,基于神经影像的AD复杂脑网络研究成为可能[7]。
综上,复杂网络分析技术与脑神经成像技术的结合为AD的早期诊断和明确其病理机制提供了全新的思路[8-9]。笔者将基于图论复杂脑网络的相关概念,对其在AD领域的应用进展和未来发展趋势等方面进行综述。
-
AD复杂脑网络研究主要基于sMRI、fMRI、PET等神经影像(表1)。He等[16]基于sMRI影像的结构性脑网络研究结果表明,AD患者较正常对照组呈现:(1)脑网络全局整合能力下降,平均路径长度更长,全局效率降低;(2)网络局部处理能力增加,如网络具有更高的聚类系数。部分研究者将这些结果解释为AD患者脑内神经元信息分离和整合平衡发生改变,导致AD患者的脑网络倾向于向规则网络(更高聚类系数,更长路径长度)方向演变,并进一步证明了这些网络参数与认知变量和记忆表现之间具有相关性[26]。
文献 成像模态 AD患者脑结构主要发现 He等[16] sMRI 聚类系数增加,平均最短路径长度增加,小世界属性减少 Supekar等[17] fMRI 聚类系数在全局水平(全脑)和局部水平(双侧海马)减少 Zhao等[18] BOLD-fMRI 聚类系数和路径长度增加 Liu等[19] fMRI 与轻度AD患者相比,重度AD患者的fMRI振幅降低,功能连接强度减少,特别是远距离连接减少 Hahn等[20]、Bernard等[21]、Dai等[22] fMRI DMN连接减少,远距离连接减少 Huang等[23] FDG PET 颞叶脑区内的功能连接减少,额叶内和顶叶与枕叶之间的功能连接减少 Titov等[24] FDG PET 额叶和顶叶区域存在较大的功能连接减少 Duan等[25] PiB PET 小世界属性增加,集聚系数和最短路径长度减少 注:表中,AD:阿尔兹海默症;sMRI:结构磁共振成像;fMRI:功能性磁共振成像;BOLD:血氧水平依赖;FDG:氟脱氧葡萄糖;PET:正电子发射断层显像术;PiB:匹斯堡复合物;DMN:默认模式网络。 表 1 AD患者的神经影像脑网络研究
Table 1. Brain network studies for AD subjects based on neuroimaging techniques
-
Supekar等[17]基于fMRI影像的功能性脑网络研究结果表明,AD患者较正常对照组呈现:(1)脑网络路径长度显著增加,即AD患者大脑加工效率明显下降,而且该网络参数与患者疾病量表显著相关[27];(2)AD患者脑网络局部连接下降[28],如默认模式网络(静息状态下相互联系、维持健康代谢活动的若干脑区组成的网络)的连接下降。此外,虽然研究发现AD患者脑功能网络局部功能连接下降,但由于大脑具有可塑性,因此AD患者脑功能网络全局功能连接仍保持稳定[29];(3) AD患者脑网络远距离连接丢失严重[28, 30]。研究结果同样表明,中心节点在AD的形成中起着关键性作用。例如,Buckner等[31]发现,AD患者在静息态fMRI网络中心节点脑区出现较高淀粉样β-淀粉样蛋白沉积,这表明AD患者的中心节点会受到疾病选择性攻击。部分研究结果同样表明,AD患者脑网络的中心节点连接减少,特别是在额叶和颞叶区域[32]。中心节点的改变在一定程度上解释了AD患者脑网络中观察到的较长路径长度和较低局部效率。总之,这些观察结果与AD患者脑网络小世界属性减弱和远距离连接的丢失是一致的[18, 33]。
-
虽然基于PET影像的AD复杂脑网络研究起步较晚,但近年来发展迅速。基于FDG PET影像的网络分析结果表明:使用稀疏逆协方差方法可以构建代谢脑网络,通过与正常组的对比发现,AD患者在颞叶脑区内的功能连接减少,且额叶内和顶叶与枕叶之间的功能连接也有减少[23-24]。Duan等[25]基于匹斯堡复合物PET影像的脑网络分析揭示,AD患者的代谢脑网络的小世界属性增强,集聚系数降低,最短路径长度明显缩短。此外,基于PET/MRI多模态影像的复杂网络分析也提高了AD疾病的早期预测能力[34]。
基于神经影像的复杂脑网络技术用于阿尔兹海默症的研究进展
Advances in the study of complex brain network based on neuroimaging in Alzheimer's disease
-
摘要: 近年来,人工智能技术被广泛地应用于阿尔兹海默症(AD)的计算机辅助诊断和疾病机理研究,基于图论的复杂网络分析技术是其中一种常见的数据挖掘方法,通过结构磁共振、功能磁共振和PET等神经成像手段获取多模态脑影像信息,结合复杂网络分析方法发现,AD患者大脑结构/功能网络存在拓扑异常改变。这一发现为AD的早期诊断和机理研究提供了新思路。笔者综述了复杂网络分析技术在AD脑结构和功能影像中的临床应用现状,并对其发展趋势进行了展望。Abstract: Artificial intelligence techniques have been widely applied in computer-aided diagnosis and disease mechanism studies for Alzheimer's disease (AD). Graph-based complex network analysis is one of the common data mining methods. A combination of complex network analysis technology and multimodal brain imaging information from neuroimaging methods, such as structural magnetic resonance imaging, functional magnetic resonance imaging, and positron emission computer imaging, could identify the abnormal changes of topological properties in brain structure and functional networks. This result provided new ideas on achieving early diagnosis and mechanism research in patients with AD. In this paper, the clinical application of complex network analysis method in structure and functional AD brain imaging was discussed, and its development trend was prospected.
-
Key words:
- Artificial intelligence /
- Alzheimer's disease /
- Brain networks /
- Neuroimaging
-
表 1 AD患者的神经影像脑网络研究
Table 1. Brain network studies for AD subjects based on neuroimaging techniques
文献 成像模态 AD患者脑结构主要发现 He等[16] sMRI 聚类系数增加,平均最短路径长度增加,小世界属性减少 Supekar等[17] fMRI 聚类系数在全局水平(全脑)和局部水平(双侧海马)减少 Zhao等[18] BOLD-fMRI 聚类系数和路径长度增加 Liu等[19] fMRI 与轻度AD患者相比,重度AD患者的fMRI振幅降低,功能连接强度减少,特别是远距离连接减少 Hahn等[20]、Bernard等[21]、Dai等[22] fMRI DMN连接减少,远距离连接减少 Huang等[23] FDG PET 颞叶脑区内的功能连接减少,额叶内和顶叶与枕叶之间的功能连接减少 Titov等[24] FDG PET 额叶和顶叶区域存在较大的功能连接减少 Duan等[25] PiB PET 小世界属性增加,集聚系数和最短路径长度减少 注:表中,AD:阿尔兹海默症;sMRI:结构磁共振成像;fMRI:功能性磁共振成像;BOLD:血氧水平依赖;FDG:氟脱氧葡萄糖;PET:正电子发射断层显像术;PiB:匹斯堡复合物;DMN:默认模式网络。 -
[1] Alzheimer's Association. 2017 Alzheimer's disease facts and figures[J]. Alzheimer's Dement, 2017, 13(4): 325−373. DOI: 10.1016/j.jalz.2017.02.001. [2] Scheltens P, Blennow K, Breteler MMB, et al. Alzheimer's disease[J]. Lancet, 2016, 388(10043): 505−517. DOI: 10.1016/S0140−6736(15)01124−1. [3] Lista S, Hampel H. Synaptic degeneration and neurogranin in the pathophysiology of Alzheimer's disease[J]. Expert Rev Neurother, 2017, 17(1): 47−57. DOI: 10.1080/14737175.2016.1204234. [4] Heneka MT, Carson MJ, El Khoury J, et al. Neuroinflammation in Alzheimer's disease[J]. Lancet Neurol, 2015, 14(4): 388−405. DOI: 10.1016/S1474−4422(15)70016−5. [5] De La Monte SM, Tong M. Brain metabolic dysfunction at the core of Alzheimer's disease[J]. Biochem Pharmacol, 2014, 88(4): 548−559. DOI: 10.1016/j.bcp.2013.12.012. [6] Jagust W. Imaging the evolution and pathophysiology of Alzheimer disease[J]. Nat Rev Neurosci, 2018, 19(11): 687−700. DOI: 10.1038/s41583−018−0067−3. [7] Stam CJ. Modern network science of neurological disorders[J]. Nat Rev Neurosci, 2014, 15(10): 683−695. DOI: 10.1038/nrn3801. [8] Bullmore E, Sporns O. The economy of brain network organization[J]. Nat Rev Neurosci, 2012, 13(5): 336−349. DOI: 10.1038/nrn3214. [9] Sporns O. Network attributes for segregation and integration in the human brain[J]. Curr Opin Neurobiol, 2013, 23(2): 162−171. DOI: 10.1016/j.conb.2012.11.015. [10] Watts DJ, Strogatz SH. Collective dynamics of 'small-world' networks[J]. Nature, 1998, 393(6684): 440−442. DOI: 10.1038/30918. [11] Oh SW, Harris JA, Ng L, et al. A mesoscale connectome of the mouse brain[J]. Nature, 2014, 508(7495): 207−214. DOI: 10.1038/nature13186. [12] Harris NG, Verley DR, Gutman BA, et al. Disconnection and hyper-connectivity underlie reorganization after TBI: A rodent functional connectomic analysis[J]. Exp Neurol, 2016, 277: 124−138. DOI: 10.1016/j.expneurol.2015.12.020. [13] Albert R, Barabási AL. Statistical mechanics of complex networks[J]. Rev Mod Phys, 2002, 74(1): 47−91. DOI: 10.1103/RevModPhys.74.47. [14] Giusti C, Ghrist R, Bassett DS. Two's company, three (or more) is a simplex: algebraic-topological tools for understanding higher-order structure in neural data[J]. J Comput Neurosci, 2016, 41(1): 1−14. DOI: 10.1007/s10827−016−0608−6. [15] Calhoun VD, Miller R, Pearlson G, et al. The Chronnectome: Time-Varying Connectivity Networks as the Next Frontier in fMRI Data Discovery[J]. Neuron, 2014, 84(2): 262−274. DOI: 10.1016/j.neuron.2014.10.015. [16] 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, 28(18): 4756−4766. DOI: 10.1523/JNEUROSCI.0141−08.2008. [17] Supekar K, Menon V, Rubin D, et al. Network Analysis of Intrinsic Functional Brain Connectivity in Alzheimer's Disease [J/OL]. PLoS Comput Biol, 2008, 4(6): e1000100[2019-11-08]. http://dx.doi.org/10.1371/journal.pcbi.1000100. DOI:10.1371/journal.pcbi.1000100. [18] Zhao QH, Lu H, Metmer H, et al. Evaluating functional connectivity of executive control network and frontoparietal network in Alzheimer's disease[J]. Brain Res, 2018, 1678: 262−272. DOI: 10.1016/j.brainres.2017.10.025. [19] Liu Y, Yu CS, Zhang XQ, et al. Impaired Long Distance Functional Connectivity and Weighted Network Architecture in Alzheimer's Disease[J]. Cereb Cortex, 2014, 24(6): 1422−1435. DOI: 10.1093/cercor/bhs410. [20] Hahn K, Myers N, Prigarin S, et al. Selectively and progressively disrupted structural connectivity of functional brain networks in Alzheimer's disease — revealed by a novel framework to analyze edge distributions of networks detecting disruptions with strong statistical evidence[J]. Neuroimage, 2013, 81: 96−109. DOI: 10.1016/j.neuroimage.2013.05.011. [21] Bernard C, Dilharreguy B, Helmer C, et al. PCC characteristics at rest in 10-year memory decliners[J]. Neurobiol Aging, 2015, 36(10): 2812−2820. DOI: 10.1016/j.neurobiolaging.2015.07.002. [22] Dai ZJ, Yan CG, Li KC, et al. Identifying and Mapping Connectivity Patterns of Brain Network Hubs in Alzheimer's Disease[J]. Cereb Cortex, 2015, 25(10): 3723−3742. DOI: 10.1093/cercor/bhu246. [23] Huang S, Li J, Sun L, et al. Learning brain connectivity of Alzheimer's disease by sparse inverse covariance estimation[J]. NeuroImage, 2010, 50(3): 935−949. DOI: 10.1016/j.neuroimage.2009.12.120. [24] Titov D, Diehl-Schmid J, Shi KY, et al. Metabolic connectivity for differential diagnosis of dementing disorders[J]. J Cereb Blood Flow Metab, 2017, 37(1): 252−262. DOI: 10.1177/0271678x15622465. [25] Duan HQ, Jiang JH, Xu J, et al. Differences in Aβ brain networks in Alzheimer's disease and healthy controls[J]. Brain Res, 2017, 1655: 77−89. DOI: 10.1016/j.brainres.2016.11.019. [26] 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(50): 16876−16885. DOI: 10.1523/JNEUROSCI.4136−10.2010. [27] Stam CJ, Jones BF, Nolte G, et al. Small-World Networks and Functional Connectivity in Alzheimer's Disease[J]. Cereb Cortex, 2006, 17(1): 92−99. DOI: 10.1093/cercor/bhj127. [28] Xue C, Yuan B, Yue Y, et al. Distinct Disruptive Patterns of Default Mode Subnetwork Connectivity Across the Spectrum of Preclinical Alzheimer's Disease[J]. Front Aging Neurosci, 2019, 11: 307. DOI: 10.3389/fnagi.2019.00307. [29] Dillen KNH, Jacobs HIL, Kukolja J, et al. Aberrant functional connectivity differentiates retrosplenial cortex from posterior cingulate cortex in prodromal Alzheimer's disease[J]. Neurobiol Aging, 2016, 44: 114−126. DOI: 10.1016/j.neurobiolaging.2016.04.010. [30] Yan TY, Wang WH, Yang L, et al. Rich club disturbances of the human connectome from subjective cognitive decline to Alzheimer's disease[J]. Theranostics, 2018, 8(12): 3237−3255. DOI: 10.7150/thno.23772. [31] Buckner RL, Sepulcre J, Talukdar T, et al. Cortical Hubs Revealed by Intrinsic Functional Connectivity: Mapping, Assessment of Stability, and Relation to Alzheimer's Disease[J]. J Neurosci, 2009, 29(6): 1860−1873. DOI: 10.1523/JNEUROSCI.5062−08.2009. [32] Filippi M, Van Den Heuvel MP, Fornito A, et al. Assessment of system dysfunction in the brain through MRI-based connectomics[J]. Lancet Neurol, 2013, 12(12): 1189−1199. DOI: 10.1016/S1474−4422(13)70144−3. [33] Evans A, He Y, Chen Z. P1-241: Structural insights into aberrant topological patterns of large-scale cortical networks in Alzheimer's disease[J]. Alzheimer's Dement, 2008, 4(4S): T284−T285. DOI: 10.1016/j.jalz.2008.05.830. [34] Zhou HC, Jiang JH, Lu JY, et al. Dual-Model Radiomic Biomarkers Predict Development of Mild Cognitive Impairment Progression to Alzheimer's Disease[J]. Front Neurosci, 2019, 12: 1045. DOI: 10.3389/fnins.2018.01045. [35] Zeidán-Chuliá F, De Oliveira BH, Salmina AB, et al. Altered expression of Alzheimer's disease-related genes in the cerebellum of autistic patients: a model for disrupted brain connectome and therapy[J/OL]. Cell Death Dis, 2014, 5(5): e1250[2019-11-08]. https://www.nature.com/articles/cddis2014227. DOI: 10.1038/cddis.2014.227. [36] Solé-Padullés C, Bartrés-Faz D, Junqué C, et al. Brain structure and function related to cognitive reserve variables in normal aging, mild cognitive impairment and Alzheimer's disease[J]. Neurobiol Aging, 2009, 30(7): 1114−1124. DOI: 10.1016/j.neurobiolaging.2007.10.008. [37] Kivelä M, Arenas A, Barthelemy M, et al. Multilayer networks[J]. J Complex Networks, 2014, 2(3): 203−271. DOI: 10.1093/comnet/cnu016. [38] Linnman C, Catana C, Petkov MP, et al. Molecular and functional PET-fMRI measures of placebo analgesia in episodic migraine: Preliminary findings[J]. Neuroimage Clin, 2018, 17: 680−690. DOI: 10.1016/j.nicl.2017.11.011. [39] Marchitelli R, Aiello M, Cachia A, et al. Simultaneous resting-state FDG-PET/fMRI in Alzheimer Disease: Relationship between glucose metabolism and intrinsic activity[J]. NeuroImage, 2018, 176: 246−258. DOI: 10.1016/j.neuroimage.2018.04.048. [40] Qin PM, Duncan NW, Chen DYT, et al. Vascular-metabolic and GABAergic Inhibitory Correlates of Neural Variability Modulation. A Combined fMRI and PET Study[J]. Neuroscience, 2018, 379: 142−151. DOI: 10.1016/j.neuroscience.2018.02.041.