Dataset Information

Using deep Siamese neural networks for detection of brain asymmetries associated with Alzheimer's Disease and Mild Cognitive Impairment.

ABSTRACT: In recent studies, neuroanatomical volume and shape asymmetries have been seen during the course of Alzheimer's Disease (AD) and could potentially be used as preclinical imaging biomarkers for the prediction of Mild Cognitive Impairment (MCI) and AD dementia. In this study, a deep learning framework utilizing Siamese neural networks trained on paired lateral inter-hemispheric regions is used to harness the discriminative power of whole-brain volumetric asymmetry. The method uses the MRICloud pipeline to yield low-dimensional volumetric features of pre-defined atlas brain structures, and a novel non-linear kernel trick to normalize these features to reduce batch effects across datasets and populations. By working with the low-dimensional features, Siamese networks were shown to yield comparable performance to studies that utilize whole-brain MR images, with the advantage of reduced complexity and computational time, while preserving the biological information density. Experimental results also show that Siamese networks perform better in certain metrics by explicitly encoding the asymmetry in brain volumes, compared to traditional prediction methods that do not use the asymmetry, on the ADNI and BIOCARD datasets.

SUBMITTER: Liu CF

PROVIDER: S-EPMC6874905 | biostudies-literature | 2019 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Using deep Siamese neural networks for detection of brain asymmetries associated with Alzheimer's Disease and Mild Cognitive Impairment.

Liu Chin-Fu CF Padhy Shreyas S Ramachandran Sandhya S Wang Victor X VX Efimov Andrew A Bernal Alonso A Shi Linyuan L Vaillant Marc M Ratnanather J Tilak JT Faria Andreia V AV Caffo Brian B Albert Marilyn M Miller Michael I MI

Magnetic resonance imaging 20190715

In recent studies, neuroanatomical volume and shape asymmetries have been seen during the course of Alzheimer's Disease (AD) and could potentially be used as preclinical imaging biomarkers for the prediction of Mild Cognitive Impairment (MCI) and AD dementia. In this study, a deep learning framework utilizing Siamese neural networks trained on paired lateral inter-hemispheric regions is used to harness the discriminative power of whole-brain volumetric asymmetry. The method uses the MRICloud pip ...[more]

PMID: 31319126

Similar Datasets

Project description:Recent research has implicated the cerebellum in Alzheimer's disease (AD), and cerebrocerebellar network connectivity is emerging as a possible contributor to symptom severity. The cerebellar dentate nucleus (DN) has parallel motor and non-motor sub-regions that project to motor and frontal regions of the cerebral cortex, respectively. These distinct dentato-cortical networks have been delineated in the non-human primate and human brain. Importantly, cerebellar regions prone to atrophy in AD are functionally connected to atrophied regions of the cerebral cortex, suggesting that dysfunction perhaps occurs at a network level. Investigating functional connectivity (FC) alterations of the DN is a crucial step in understanding the cerebellum in AD and in mild cognitive impairment (MCI). Inclusion of this latter group stands to provide insights into cerebellar contributions prior to diagnosis of AD. The present study investigated FC differences in dorsal (dDN) and ventral (vDN) DN networks in MCI and AD relative to cognitively normal participants (CN) and relationships between FC and behavior. Our results showed patterns indicating both higher and lower functional connectivity in both dDN and vDN in AD compared to CN. However, connectivity in the AD group was lower when compared to MCI. We argue that these findings suggest that the patterns of higher FC in AD may act as a compensatory mechanism. Additionally, we found associations between the individual networks and behavior. There were significant interactions between dDN connectivity and motor symptoms. However, both DN seeds were associated with cognitive task performance. Together, these results indicate that cerebellar DN networks are impacted in AD, and this may impact behavior. In concert with the growing body of literature implicating the cerebellum in AD, our work further underscores the importance of investigations of this region. We speculate that much like in psychiatric diseases such as schizophrenia, cerebellar dysfunction results in negative impacts on thought and the organization therein. Further, this is consistent with recent arguments that the cerebellum provides crucial scaffolding for cognitive function in aging. Together, our findings stand to inform future clinical work in the diagnosis and understanding of this disease.

Dataset Information

Using deep Siamese neural networks for detection of brain asymmetries associated with Alzheimer's Disease and Mild Cognitive Impairment.

Publications

Using deep Siamese neural networks for detection of brain asymmetries associated with Alzheimer's Disease and Mild Cognitive Impairment.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets