Classification of Alzheimer's disease using ensemble of deep neural networks trained through transfer learning

AbstractDeep neural networks (DNNs) capture complex relationships among variables, however, because they require copious samples, their potential has yet to be fully tapped for understanding relationships between gene expression and human phenotypes. Here we introduce an analysis framework, namely MD-AD (Multi-task Deep learning for Alzheimer’s Disease neuropathology), which leverages an unexpected synergy between DNNs and multi-cohort settings. In these settings, true joint analysis can be stymied using conventional statistical methods, which require “harmonized” phenotypes and tend to capture cohort-level variations, obscuring subtler true disease signals. Instead, MD-AD incorporates related phenotypes sparsely measured across cohorts, and learns interactions between genes and phenotypes not discovered using linear models, identifying subtler signals than cohort-level variations which can be uniquely recapitulated in animal models and across tissues. We show that MD-AD exploits sex-specific relationships between microglial immune response and neuropathology, providing a nuanced context for the association between inflammatory genes and Alzheimer’s Disease.

Download Full-text

Predicting late-onset Alzheimer’s disease from genomic data using deep neural networks

10.1101/629402 ◽

2019 ◽

Author(s):

Javier de Velasco Oriol ◽

Edgar E. Vallejo ◽

Karol Estrada ◽

Keyword(s):

Alzheimer’S Disease ◽

Neural Networks ◽

Alzheimer's Disease ◽

Genetic Variants ◽

Deep Neural Networks ◽

Late Onset ◽

Association Studies ◽

Genome Wide Association Studies ◽

Clinical Markers ◽

Genome Wide

AbstractAlzheimer’s disease (AD) is the leading form of dementia. Over 25 million cases have been estimated worldwide and this number is predicted to increase two-fold every 20 years. Even though there is a variety of clinical markers available for the diagnosis of AD, the accurate and timely diagnosis of this disease remains elusive. Recently, over a dozen of genetic variants predisposing to the disease have been identified by genome-wide association studies. However, these genetic variants only explain a small fraction of the estimated genetic component of the disease. Therefore, useful predictions of AD from genetic data could not rely on these markers exclusively as they are not sufficiently informative predictors. In this study, we propose the use of deep neural networks for the prediction of late-onset Alzheimer’s disease from a large number of genetic variants. Experimental results indicate that the proposed model holds promise to produce useful predictions for clinical diagnosis of AD.

Download Full-text

Deep Learning in Brain MRI: Effect of Data Leakage Due to Slice-level Split Using 2D Convolutional Neural Networks

10.21203/rs.3.rs-464091/v1 ◽

2021 ◽

Author(s):

Ekin Yagis ◽

Selamawet Workalemahu Atnafu ◽

Alba García Seco de Herrera ◽

Chiara Marzi ◽

Marco Giannelli ◽

...

Keyword(s):

Alzheimer’S Disease ◽

Neural Networks ◽

Alzheimer's Disease ◽

Convolutional Neural Networks ◽

De Novo ◽

Neurological Diseases ◽

Brain Mri ◽

Progression Markers ◽

Magnetic Resonance Imaging Mri

Abstract In recent years, 2D convolutional neural networks (CNNs) have been extensively used for the diagnosis of neurological diseases from magnetic resonance imaging (MRI) data due to their potential to discern subtle and intricate patterns. Despite the high performances reported in numerous studies, developing CNN models with good generalization abilities is still a challenging task due to possible data leakage introduced during cross-validation (CV). In this study, we quantitatively assessed the effect of a data leakage caused by 3D MRI data splitting based on a 2D slice-level using three 2D CNN models for the classification of patients with Alzheimer’s disease (AD) and Parkinson’s disease (PD). Our experiments showed that slice-level CV erroneously boosted the average slice level accuracy on the test set by 30% on Open Access Series of Imaging Studies (OASIS), 29% on Alzheimer’s Disease Neuroimaging Initiative (ADNI), 48% on Parkinson's Progression Markers Initiative (PPMI) and 55% on a local de-novo PD Versilia dataset. Further tests on a randomly labeled OASIS-derived dataset produced about 96% of (erroneous) accuracy (slice-level split) and 50% accuracy (subject-level split), as expected from a randomized experiment. Overall, the extent of the effect of an erroneous slice-based CV is severe, especially for small datasets.

Download Full-text