Dataset Information

Deep Learning of Static and Dynamic Brain Functional Networks for Early MCI Detection.

ABSTRACT: While convolutional neural network (CNN) has been demonstrating powerful ability to learn hierarchical spatial features from medical images, it is still difficult to apply it directly to resting-state functional MRI (rs-fMRI) and the derived brain functional networks (BFNs). We propose a novel CNN framework to simultaneously learn embedded features from BFNs for brain disease diagnosis. Since BFNs can be built by considering both static and dynamic functional connectivity (FC), we first decompose rs-fMRI into multiple static BFNs with modified independent component analysis. Then, the voxel-wise variability in dynamic FC is used to quantify BFN dynamics. A set of paired 3D images representing static/dynamic BFNs can be fed into 3D CNNs, from which we can hierarchically and simultaneously learn static/dynamic BFN features. As a result, the dynamic BFN features can complement static BFN features and, at the meantime, different BFNs can help each other toward a joint and better classification. We validate our method with a publicly accessible, large cohort of rs-fMRI dataset in early-stage mild cognitive impairment (eMCI) diagnosis, which is one of the most challenging problems to the clinicians. By comparing with a conventional method, our method shows significant diagnostic performance improvement by almost 10%. This result demonstrates the effectiveness of deep learning in preclinical Alzheimer's disease diagnosis, based on the complex and high-dimensional voxel-wise spatiotemporal patterns of the resting-state brain functional connectomics. The framework provides a new but intuitive way to fully exploit deeply embedded diagnostic features from rs-fMRI for a better-individualized diagnosis of various neurological diseases.

SUBMITTER: Kam TE

PROVIDER: S-EPMC7122732 | biostudies-literature | 2020 Feb

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Deep Learning of Static and Dynamic Brain Functional Networks for Early MCI Detection.

Kam Tae-Eui TE Zhang Han H Zhang Han H Jiao Zhicheng Z Shen Dinggang D

IEEE transactions on medical imaging 20190717 2

While convolutional neural network (CNN) has been demonstrating powerful ability to learn hierarchical spatial features from medical images, it is still difficult to apply it directly to resting-state functional MRI (rs-fMRI) and the derived brain functional networks (BFNs). We propose a novel CNN framework to simultaneously learn embedded features from BFNs for brain disease diagnosis. Since BFNs can be built by considering both static and dynamic functional connectivity (FC), we first decompos ...[more]

PMID: 31329111

Similar Datasets

Project description:IntroductionPost-stroke aphasia (PSA) is a language disorder caused by left hemisphere stroke. Electroacupuncture (EA) is a minimally invasive therapeutic option for PSA treatment. Tongli (HT5) and Xuanzhong (GB39), two important language-associated acupoints, are frequently used in the rehabilitation of patients with PSA. Preliminary evidence indicated functional activation in distributed cortical areas upon HT5 and GB39 stimulation. However, research on the modulation of dynamic and static functional connectivity (FC) in the brain by EA in PSA is lacking.MethodThis study aimed to investigate the PSA-related effects of EA stimulation at HT5 and GB39 on neural processing. Thirty-five participants were recruited, including 19 patients with PSA and 16 healthy controls (HCs). The BOLD signal was analyzed by static independent component analysis, generalized psychophysiological interactions, and dynamic independent component analysis, considering variables such as age, sex, and years of education.ResultsThe results revealed that PSA showed activated clusters in the left putamen, left postcentral gyrus (PostCG), and left angular gyrus in the salience network (SN) compared to the HC group. The interaction effect on temporal properties of networks showed higher variability of SN (F = 2.23, positive false discovery rate [pFDR] = 0.017). The interaction effect on static FC showed increased functional coupling between the right calcarine and right lingual gyrus (F = 3.16, pFDR = 0.043). For the dynamic FC, at the region level, the interaction effect showed lower variability and higher frequencies of circuit 3, with the strongest connections between the supramarginal gyrus and posterior cingulum (F = 5.42, pFDR = 0.03), middle cingulum and PostCG (F = 5.27, pFDR = 0.036), and triangle inferior frontal and lingual gyrus (F = 5.57, pFDR = 0.026). At the network level, the interaction effect showed higher variability in occipital network-language network (LN) and cerebellar network (CN) coupling, with stronger connections between the LN and CN (F = 4.29, pFDR = 0.042). Dynamic FC values between the triangle inferior frontal and lingual gyri were anticorrelated with transcribing, describing, and dictating scores in the Chinese Rehabilitation Research Center for Chinese Standard Aphasia Examination.DiscussionThese findings suggest that EA stimulation may improve language function, as it significantly modulated the nodes of regions/networks involved in the LN, SN, CN, occipital cortex, somatosensory regions, and cerebral limbic system.

Dataset Information

Deep Learning of Static and Dynamic Brain Functional Networks for Early MCI Detection.

Publications

Deep Learning of Static and Dynamic Brain Functional Networks for Early MCI Detection.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets