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

Hahn, Klaus; Myers, Norman; Prigarin, S. M.; Rodenacker, Karsten; Kurz, Alexander; Förstl, Hans; Zimmer, Claus; Wohlschläger, Afra M.; Sorg, Christian

doi:10.1016/j.neuroimage.2013.05.011

Cited by 82 publications

(58 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Suppose i ∈ [1,2,3,4,5] denotes the run index of cross validation process. The whole dataset S is divided into five folds [S 1 , S 2 , S 3 , S 4 , S 5 ].…”

Section: K-fold Cross Validationmentioning

confidence: 99%

See 1 more Smart Citation

Classification of Alzheimer Disease Based on Structural Magnetic Resonance Imaging by Kernel Support Vector Machine Decision Tree

Zhang¹,

Wang²,

Dong³

2014

PIER

192

View full text Add to dashboard Cite

Abstract-In this paper we proposed a novel classification system to distinguish among elderly subjects with Alzheimer's disease (AD), mild cognitive impairment (MCI), and normal controls (NC). The method employed the magnetic resonance imaging (MRI) data of 178 subjects consisting of 97 NCs, 57 MCIs, and 24 ADs. First, all these three dimensional (3D) MRI images were preprocessed with atlas-registered normalization. Then, gray matter images were extracted and the 3D images were under-sampled. Afterwards, principle component analysis was applied for feature extraction. In total, 20 principal components (PC) were extracted from 3D MRI data using singular value decomposition (SVD) algorithm, and 2 PCs were extracted from additional information (consisting of demographics, clinical examination, and derived anatomic volumes) using alternating least squares (ALS). On the basic of the 22 features, we constructed a kernel support vector machine decision tree (kSVM-DT). The error penalty parameter C and kernel parameter σ were determined by Particle Swarm Optimization (PSO). The weights ω and biases b were still obtained by quadratic programming method. 5-fold cross validation was employed to obtain the out-of-sample estimate. The results show that the proposed kSVM-DT achieves 80% classification accuracy, better than 74% of the method without kernel. Besides, the PSO exceeds the random selection method in choosing the parameters of the classifier. The computation time to predict a new patient is only 0.022 s.

show abstract

“…Suppose i ∈ [1,2,3,4,5] denotes the run index of cross validation process. The whole dataset S is divided into five folds [S 1 , S 2 , S 3 , S 4 , S 5 ].…”

Section: K-fold Cross Validationmentioning

confidence: 99%

“…There is no cure for this disease, which is progressively worsening and eventually leads to death [1]. In 2006, there were 26.6 million sufferers worldwide, and AD is predicted to affect 1 in 85 people globally by 2050, and at least 43% of prevalent cases need a high level of care [2].…”

Section: Introductionmentioning

confidence: 99%

Classification of Alzheimer Disease Based on Structural Magnetic Resonance Imaging by Kernel Support Vector Machine Decision Tree

Zhang¹,

Wang²,

Dong³

2014

PIER

192

View full text Add to dashboard Cite

show abstract

“…These overlaps were found especially for the default mode network (13,16), one of the most prominent and consistent ICNs associated with the processing of internally focused tasks (17), as well as also for other frontoparietal ICNs (18). Importantly, it also has been demonstrated that AD and MCI patients show disruptions of functional and structural connectivity within ICNs (19,20) and that even elderly nondemented 11 C-PiB-positive subjects show associations between hypometabolism, 11 C-PiB-uptake, and connectivity disruption (21,22).…”

mentioning

confidence: 94%

Prefrontal Hypometabolism in Alzheimer Disease Is Related to Longitudinal Amyloid Accumulation in Remote Brain Regions

et al. 2015

Self Cite

View full text Add to dashboard Cite

In PET studies of patients with Alzheimer disease (AD), prominent hypometabolism can occur in brain regions without major amyloid load. These hypometabolism-only (HO) areas may not be explained easily as a consequence of local amyloid toxicity. The aim of this longitudinal multimodal imaging study was the investigation of locoregional and remote relationships between metabolism in HO areas and longitudinal amyloid increase in functionally connected brain areas, with a particular focus on intrinsic functional connectivity as a relevant linking mechanism between pathology and dysfunction. Methods: Fifteen AD patients underwent longitudinal examinations with 11 C-Pittsburgh compound B ( 11 C-PiB) and 18 F-FDG PET (mean follow-up period, 2 y). The peak HO region was identified by the subtraction of equally thresholded statistical T maps (hypometabolism minus amyloid burden), resulting from voxel-based statistical parametric mapping group comparisons between the AD patients and 15 healthy controls. Then functionally connected and nonconnected brain networks were identified by means of seedbased intrinsic functional connectivity analysis of the resting-state functional MRI data of healthy controls. Finally, network-based, regionof-interest-based, and voxel-based correlations were calculated between longitudinal changes of normalized 11 C-PiB binding and 18 F-FDG metabolism. Results: Positive voxel-based and region-ofinterest-based correlations were demonstrated between longitudinal 11 C-PiB increases in the HO-connected network, encompassing bilateral temporoparietal and frontal brain regions, and metabolic changes in the peak HO region as well as locoregionally within several AD-typical brain regions. Conclusion: Our results indicate that in AD amyloid accumulation in remote but functionally connected brain regions may significantly contribute to longitudinally evolving hypometabolism in brain regions not strongly affected by local amyloid pathology, supporting the amyloid-and networkdegeneration hypothesis. Al zheimer disease (AD) is a progressive neurodegenerative disorder (1). Current hypotheses about the pathogenesis assume that the aggregation of b-amyloid peptides plays an important role and represents an early event of this disease by leading to synaptic dysfunction and synaptic loss, finally causing neuronal cell death and cognitive decline (2,3). The understanding of causal relations and chronologic sequences of amyloid deposition and neuronal dysfunction in the development of AD is important for the classification of disease severity and for the evaluation of new treatment approaches. The current study aims at such sequences of patterns of both amyloid plaque deposition and neural dysfunction, with particular focus on intrinsic functional connectivity as a relevant linking mechanism between pathology and dysfunction.With noninvasive PET biomarkers, such as Pittsburgh compound B ( 11 C-PiB) and 18 F-FDG, AD-related pathologies, namely fibrillary amyloid burden and hypometabolism, as well as their longitudina...

show abstract

“…Application of the latter technique to datasets including subjects ranging from normally aging individuals to patients with mild cognitive impairment (MCI) or Alzheimer disease dementia (AD) has revealed a selective breakdown of intrinsic brain networks during the progression of AD (5,6).…”

mentioning

confidence: 99%

Progressive Disintegration of Brain Networking from Normal Aging to Alzheimer Disease: Analysis of Independent Components of ¹⁸F-FDG PET Data

et al. 2017

View full text Add to dashboard Cite

Brain connectivity has been assessed in several neurodegenerative disorders investigating the mutual correlations between predetermined regions or nodes. Selective breakdown of brain networks during progression from normal aging to Alzheimer disease dementia (AD) has also been observed. Methods: We implemented independentcomponent analysis of 18 F-FDG PET data in 5 groups of subjects with cognitive states ranging from normal aging to AD-including mild cognitive impairment (MCI) not converting or converting to ADto disclose the spatial distribution of the independent components in each cognitive state and their accuracy in discriminating the groups. Results: We could identify spatially distinct independent components in each group, with generation of local circuits increasing proportionally to the severity of the disease. AD-specific independent components first appeared in the late-MCI stage and could discriminate converting MCI and AD from nonconverting MCI with an accuracy of 83.5%. Progressive disintegration of the intrinsic networks from normal aging to MCI to AD was inversely proportional to the conversion time. Conclusion: Independent-component analysis of 18 F-FDG PET data showed a gradual disruption of functional brain connectivity with progression of cognitive decline in AD. This information might be useful as a prognostic aid for individual patients and as a surrogate biomarker in intervention trials.

show abstract

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

Cited by 82 publications

References 52 publications

Classification of Alzheimer Disease Based on Structural Magnetic Resonance Imaging by Kernel Support Vector Machine Decision Tree

Classification of Alzheimer Disease Based on Structural Magnetic Resonance Imaging by Kernel Support Vector Machine Decision Tree

Prefrontal Hypometabolism in Alzheimer Disease Is Related to Longitudinal Amyloid Accumulation in Remote Brain Regions

Progressive Disintegration of Brain Networking from Normal Aging to Alzheimer Disease: Analysis of Independent Components of ¹⁸F-FDG PET Data

Contact Info

Product

Resources

About

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

Cited by 82 publications

References 52 publications

Classification of Alzheimer Disease Based on Structural Magnetic Resonance Imaging by Kernel Support Vector Machine Decision Tree

Classification of Alzheimer Disease Based on Structural Magnetic Resonance Imaging by Kernel Support Vector Machine Decision Tree

Prefrontal Hypometabolism in Alzheimer Disease Is Related to Longitudinal Amyloid Accumulation in Remote Brain Regions

Progressive Disintegration of Brain Networking from Normal Aging to Alzheimer Disease: Analysis of Independent Components of 18F-FDG PET Data

Contact Info

Product

Resources

About

Progressive Disintegration of Brain Networking from Normal Aging to Alzheimer Disease: Analysis of Independent Components of ¹⁸F-FDG PET Data