脑影像智能分析与脑疾病早期诊断ppt课件.pptx

,脑影像智能分析与脑疾病 早期诊断,张道强南京航空航天大学,Brain Projects,美国“脑活动图谱”计划,欧盟“人类脑计划”,中国脑计划,Brain Imaging (Neuroimaging),Neuroimaging includes the use of various techniques to either directly or indirectly image the structure or function of the brainTwo broad categoriesStructural neuroimaging deals with the structure of the brainFunctional neuroimaging is used to indirectly measure brain functions,Neuroimaging-based Classification,(S. Lemm, et al., Neuroimage, 2011),Example: Brain Decoding,(Nature Feature News, 2013),(T. Mitchell et al., Science, 2008),Recovery Movies,Outline,Summary,1,2,3,Backgrounds on Alzheimers Disease,Brain-imaging based Analysis,Brain-network based Analysis,4,History of AD,AD was first described by German psychiatrist and neuropathologist Alois Alzheimer in 1906 and was named after himThe 51 y.o. woman (Auguste Deter) cared by Dr. Alzheimer until her death in 1906. He did an autopsy, examined her brain & described the typical abnormalities of what would be called later Alzheimers Disease,What Is AD?,It is the most common form of dementiaThere is no cure for the disease, which worsens as it progresses, and eventually leads to deathMost often, AD is diagnosed in people over 65 years of ageIn 2006, there were 26.6 million sufferers worldwide, and it is predicted to affect 1 in 85 people globally by 2050,不同年代我国痴呆和AD患者的人数【 Lancet. 2013】,“三高”：患病率高、 致残率高、负担重“三低”：就诊率低、 诊断率低、治疗率低目前我国AD的患病率 【 Alzheimers & Dementia. 2013】,我国AD现状:,Celebrities with AD,Normal vs. AD Brain,In the normal brain there is a lot of healthy brain tissue in the language area. In the AD affected brain there is little in that areaThere are many differences between the two brains including the memory, sulcus, gyrus, ventricle, and language areas. In the AD brain, these are either shrunken or stretched out to unhealthy measures,Normal vs. AD Brain,Forms abnormal clumps called amyloid plaques and tangled bundles of fibers called neurofibrillary tangles in the brain,AD自画像,1967（早年）,1996（患病第2年）,1997（患病第3年）,1998,1999,2000,Normal or diseased?,(S. Crutch,et al., Lancet Neurology, 2012),Normal or diseased?,(S. Crutch,et al., Lancet Neurology, 2012),AD Progression,AD atrophy progressesStarts in the medial temporal and limbic areas Hippocampus and entorhinal cortexSubsequently spreading to parietal association areasFinally to frontal and primary cortices,AD Biomarkers,Biomarkers for early diagnosis of ADMagnetic resonance imaging (MRI)Positron emission tomography (PET)Cerebrospinal fluid (CSF)- A42, t-tau and p-tau,MRI,PET,CSF,Biomarkers,Outline,Summary,1,2,3,Backgrounds on Alzheimers Disease,Brain-imaging based Analysis,Brain-network based Analysis,4,Multimodal Classification,Motivation Several modalities of biomarkers have been proved to be sensitive to AD, or its prodromal stage, i.e., mild cognitive impairment (MCI) Different biomarkers provide complementary information, which may be useful for diagnosis of AD or MCI when used togetherQuestion: How can we effectively combine both imaging data (MRI and PET) and non-imaging data (CSF) for multi- modality based classification?,Flowchart,Template,MRIdata,PETdata,CSFdata,68,131,21,42,Featureextraction,Featureextraction,Feature selection,Calculatekernel matrix,Calculatekernel matrix,Calculatekernel matrix,SVM,optional,KKeerrnneell ccoommbbiinnaattiioonn,(D. Zhang, et al. Neuroimage, 2011),Materials,202 subjects from ADNI, including 51 AD patients, 99 MCI and 52 healthy controls43 MCI converters who had converted to AD within 18 monthsand 56 MCI non-converters who had not convertedOnly baseline data of MRI, CSF and PET are used,Results,Comparison of performance of single-modaland multimodal classification methods,(D. Zhang, et al. Neuroimage, 2011),Multi-Modal Multi-Task Learning,MotivationBesides classification, there also exist regression tasks which estimate continuous clinical scores to evaluate the stage of AD pathology and predict future progressionBoth regression and classification tasks are essentially related due to the same underlying pathologyQuestion: How can we jointly predict multiple regression and classification variables from multi-,modality data?,AD/MCI/HCMMSEADAS-Cog,(D. Zhang, D. Shen. Neuroimage, 2012),Flowchart,Template,MRIdata,PETdata,CSFdata,Featureextraction,Featureextraction,SVM(Regression/Classification),68, 131,21,42,Clinicalscores,MMSE；ADAS-CogClass Labels,Calculate kernel matrix,SVM(Regression/Classification),Multi-modelSVMKKeerrnneell combinationCalculatekkeerrnneell mmaattrriixxCCaallccuullaattee kernel matrix,MTFSMulti-task feature selectionMTFS,(D. Zhang, D. Shen. Neuroimage, 2012),Materials,ADNI Subjects186subjects (45AD, 91 MCI and 50 HCs), only baseline data, 3 modalities (MRI, CSF and PET),deviati,Experiments,Experiment 1Estimating clinical stagesMMSE, ADAS-Cog, and class label (AD/MCI/HC)Experiment 2Predicting 2-year MMSE and ADAS-Cog changes and MCI conversion,Results,Comparison of performances of different methods on Experiment 1,Results (contd),Comparison of performances of different methods on Experiment 2,Manifold Regularized Multitask Learning,(B. Jie, D. Zhang, et al. Human Brain Mapping, 2015),Multi-level Multitask Learning,(M. Wang, et al. MICCAI 2017),Longitudinal Multitask Learning,(D. Zhang, et al. PLOS ONE, 2012; B. Jie, et al. IEEE TBME 2017),Multimodal Transfer Learning,(B. Cheng, et al. Brain Imaging Neuroinformatics 2017),Multi-Atlas Classification,(M. Liu, D. Zhang, et al. Human Brain Mapping, 2015; IEEE TMI 2016; TBME 2016),Illustration of multi-center adaptation with low-rankrepresentation learning for disease diagnosis(M. Wang, D. Zhang and etc., MICCAI 2018),Multi-Center Disease Classification,Low-Rank Representation Method,(M. Wang, D. Zhang and etc., MICCAI 2018),Imaging Genetics,Association,Tree Structure Among SNPs,(a) group by gene,(b) group by linkage disequilibrium (LD) cluster,Tree-Guided Sparse Learning,TGSL,.,SNP2,SNP4,SNPi,SNP1,.,SVR,Tree Construction by Gene or LD Hierarchical Clustering,SNPsSelection,SNPk,MRISimultaneous feature selection and regr,TGSL,( X. Hao, et al. MICCAI, 2014; IEEE TCBB 2018),Multi-modality Phenotype Associations,(X. Hao, et al. PSB, 2016; Neuroinformatics 2016),Multi-SNP-Multi-QT Associations,(X. Hao, et al. Scientific Reports 2017),wr,w3=0,v3,x1u1,x2u2=0,x3u3,x4u4=0,xpupw1=0w2,.,v1=0y1,v2=0y2,y3,vqyq,.,X,Y,z1z2x3.zr,Z,Longitudinal Phenotype Associations,(X. Hao, et al. Bioinformatics 2017),Brain Decoding: Hyperalignment,Brain Decoding: Hyperalignment,Local Discriminant Hyperalignment,(Yousefnezhad & Zhang, AAAI 2017),Original Space,Common Space ( ),Subject 1,Subject S,Voxel x2,Voxel,Voxel,Voxel x2,Deep Hyperalignment,(Yousefnezhad & Zhang, NIPS 2017),Outline,Summary,1,2,3,Backgrounds on Alzheimers Disease,Brain-imaging based Analysis,Brain-network based Analysis,4,Brain Connectomics,Studies the interaction of brain functional regions at systems level,(Petra E. Vrtes, et al.,PNAS, 2012)Mapping the human brain is one of the great scientific challenges of the 21st century,Connectivity Analysis,(Honey et al., PNAS, 2007),Network-based Classification,Neuroimaging data,Networks construction,Feature extraction and selection,Training classifier,Network-based Classification,Motivation Brain connectivity networks have been used forclassification of AD/MCI from normal controls (NC) In conventional methods, local measures of connectivity networks are first extracted from each ROI as network features, and then concatenated into a long vector Some useful structural information of network, especiallyglobal topological information, may be lostQuestion: How can we better preserve the network topological information for more effective brain network based classification?,Topological Graph Kernel,Topology-based graph kernelThe kernel is defined on graphs, which can be used to computethe similarity of a pair of graphs,fMRI image,Regional mean time series,Connectivity network,Gray matter mask,T1,T2,TM,RFE-GK,RFE-GK,RFE-GK,Kernel combination,Kernel matrix,Kernelmatrix,Kernelmatrix,0.40.20-0.2-0.4-0.6,0.80.6,t-test,t-test,t-test,SVM Classifier,0.90.80.70.60.50.40.30.20.1,T1,0.90.80.70.60.50.40.30.20.1,0.90.80.70.60.50.40.30.20.1,T,2,TM,Thresholded connectivity networks,FFeeaattuurreeextraction,RFE-GK,RFE-GK,RFE-GK,t-test,t-test,t-test,FFeeaattuurree selection,Flowchart,(B. Jie, D. Zhang, et al., Human Brain Mapping, 2014),Classification results,ROC curve,0.10.2,1,00,0.1,0.2,0.3,0.4,10.90.80.70.60.5,0.30.40.50.60.70.80.9False positive rate(1-Specificity),True positive rate(sensitivity),ROC curve,Random VEC+RFE-LK VEC+RFE-RBFt-test Combined RFE-RBF Combined RFE-LK Combined RFE-GK Combined,Brain Sub-networks,T3(a) NC(b) MCIThresholded average connectivity sub-network basedon top selected ROIs,T2,T1,(B. Jie, et al.IEEE TBME,2014),(B. Jie, et al.IEEE TIP,2018),Sub-network Kernel Based Method,Hyper-network based Method,MotivationConventional connectivity network is usually constructed based on the pairwise correlation among brain regionsCannot reflect the useful higher-order relationship among brain regions,Question: how to character the higher-order relationship among brain regions?,Solution: Hyper-graph,V1,V3,V,2,V4,V5,V6,V7, = ( , ), = , , , , = , , , = , , = , , ,= , , = , , ,V,V7,V6,e1,5e3,V3e2,e4,V2V4,V1,Hyper-graph vs. graph. Left: a conventional graph in which two nodes are connect- edtogether by an edge. Middle: a hyper-graph in which each hyper-edge can connect morethan two nodes. Right: the incidence matrix for the hyper-graph in the middle,Flowchart,(B. Jie, D. Shen, D. Zhang, MICCAI14; Medical Image Analysis 2016),Experimental Results,Classification performances of different methods,Results,0.1,0.2,0.8,0.9,1,00,0.1,0.2,0.3,0.4,0.5,10.90.80.70.6,0.30.40.50.60.7False positive rate,True positive rate,ROC curve,CN_CC HN_HCC1 HN_HCC2 HN_HCC3Proposed,Hyperedges,(c) L. anterior cingulate gyrus (L.ACG) NCMCI,(d) L. middle cingulate gyrus (L.MCG)NCMCI,(a) L. middle frontal gyrus (L.MFG),(b) L. rectus gyrus (L.REC),Connectome Biomarkers,(C. Zu, et al. Brain Imaging and Behavior 2018),Top 10 Subnetworks,(C. Zu, et al. Brain Imaging and Behavior 2018),Ordinal Patterns Mining,Existing network descriptorsNode degreesClustering coefficientsSub-networks Limitations of previous workDesigned on un-weighted brain connectivity networksFocusonindividualbrainregionsotherthanlocal structures of brain networks,An overview of ordinal pattern based learning for brain disease diagnosis(X. Liu, et al., MICCAI 2016; D. Zhang, et al., IEEE TMI 2018),Flowchart,Illustration of the proposed ordinal patterns,Illustration,Experimental Results,Comparison of different methods in three classification tasks,(a) Top 2 ordinal patters from ADHD,(b) Top 2 ordinal patters from NC,Ordinal patterns identified in AD vs. NC classification,Identified Ordinal Patterns,Thanks for your attention!,