Dataset Information

Sociodemographic data and APOE-?4 augmentation for MRI-based detection of amnestic mild cognitive impairment using deep learning systems.

ABSTRACT: Detection and diagnosis of early and subclinical stages of Alzheimer's Disease (AD) play an essential role in the implementation of intervention and prevention strategies. Neuroimaging techniques predominantly provide insight into anatomic structure changes associated with AD. Deep learning methods have been extensively applied towards creating and evaluating models capable of differentiating between cognitively unimpaired, patients with Mild Cognitive Impairment (MCI) and AD dementia. Several published approaches apply information fusion techniques, providing ways of combining several input sources in the medical domain, which contributes to knowledge of broader and enriched quality. The aim of this paper is to fuse sociodemographic data such as age, marital status, education and gender, and genetic data (presence of an apolipoprotein E (APOE)-?4 allele) with Magnetic Resonance Imaging (MRI) scans. This enables enriched multi-modal features, that adequately represent the MRI scan visually and is adopted for creating and modeling classification systems capable of detecting amnestic MCI (aMCI). To fully utilize the potential of deep convolutional neural networks, two extra color layers denoting contrast intensified and blurred image adaptations are virtually augmented to each MRI scan, completing the Red-Green-Blue (RGB) color channels. Deep convolutional activation features (DeCAF) are extracted from the average pooling layer of the deep learning system Inception_v3. These features from the fused MRI scans are used as visual representation for the Long Short-Term Memory (LSTM) based Recurrent Neural Network (RNN) classification model. The proposed approach is evaluated on a sub-study containing 120 participants (aMCI = 61 and cognitively unimpaired = 59) of the Heinz Nixdorf Recall (HNR) Study with a baseline model accuracy of 76%. Further evaluation was conducted on the ADNI Phase 1 dataset with 624 participants (aMCI = 397 and cognitively unimpaired = 227) with a baseline model accuracy of 66.27%. Experimental results show that the proposed approach achieves 90% accuracy and 0.90 F1-Score at classification of aMCI vs. cognitively unimpaired participants on the HNR Study dataset, and 77% accuracy and 0.83 F1-Score on the ADNI dataset.

SUBMITTER: Pelka O

PROVIDER: S-EPMC7518632 | biostudies-literature | 2020

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Sociodemographic data and APOE-ε4 augmentation for MRI-based detection of amnestic mild cognitive impairment using deep learning systems.

Pelka Obioma O Friedrich Christoph M CM Nensa Felix F Mönninghoff Christoph C Bloch Louise L Jöckel Karl-Heinz KH Schramm Sara S Sanchez Hoffmann Sarah S Winkler Angela A Weimar Christian C Jokisch Martha M

PloS one 20200925 9

Detection and diagnosis of early and subclinical stages of Alzheimer's Disease (AD) play an essential role in the implementation of intervention and prevention strategies. Neuroimaging techniques predominantly provide insight into anatomic structure changes associated with AD. Deep learning methods have been extensively applied towards creating and evaluating models capable of differentiating between cognitively unimpaired, patients with Mild Cognitive Impairment (MCI) and AD dementia. Several p ...[more]

PMID: 32976486

Similar Datasets

Project description:BACKGROUND:Mild cognitive impairment (MCI) is a heterogeneous syndrome resulting from Alzheimer's disease (AD) as well as to non-AD and non-neurodegenerative conditions. A subset of patients with amnestic MCI (aMCI) present with an unusually long-lasting course, a slow rate of clinical neuropsychological progression, and evidence of focal involvement of medial temporal lobe structures. In the present study, we explored positron emission tomography (PET) and cerebrospinal fluid (CSF) biomarkers in a sample of subjects with aMCI with such clinical features in order to provide in vivo evidence to improve disease characterisation in this subgroup. METHODS:Thirty consecutive subjects with aMCI who had long-lasting memory impairment (more than 4 years from symptom onset) and a very slow rate of cognitive progression were included. All subjects underwent fluorodeoxyglucose-positron emission tomography (FDG-PET) metabolic imaging. A measure of cerebral amyloid load, by PET and/or CSF, was obtained in 26 of 30 subjects. The mean clinical follow-up was 58.3?±?10.1 months. RESULTS:No patient progressed to dementia during the follow-up. The typical AD FDG-PET pattern of temporoparietal hypometabolism was not present in any of the subjects. In contrast, a selective medial temporal lobe hypometabolism was present in all subjects, with an extension to frontolimbic regions in some subjects. PET imaging showed absent or low amyloid load in the majority of samples. The values were well below those reported in prodromal AD, and they were slightly elevated in only two subjects, consistent with the CSF ?-amyloid (1-42) protein values. Notably, no amyloid load was present in the hippocampal structures. CONCLUSIONS:FDG-PET and amyloid-PET together with CSF findings questioned AD pathology as a unique neuropathological substrate in this aMCI subgroup with long-lasting disease course. The possibility of alternative pathological conditions, such as argyrophilic grain disease, primary age-related tauopathy or age-related TDP-43 proteinopathy, known to spread throughout the medial temporal lobe and limbic system structures should be considered in these patients with MCI.

Dataset Information

Sociodemographic data and APOE-?4 augmentation for MRI-based detection of amnestic mild cognitive impairment using deep learning systems.

Publications

Sociodemographic data and APOE-ε4 augmentation for MRI-based detection of amnestic mild cognitive impairment using deep learning systems.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets