Structural‐MRI‐based Alzheimer's disease dementia score using 3D convolutional neural networks to achieve accurate early disease prediction: Neuroimaging / Optimal neuroimaging measures for early detection. (7th December 2020)

• Record Type: Journal Article
• Authors: Yee, Evangeline; Popuri, Karteek; Ma, Da; Wang, Lei; Beg, Mirza Faisal
• Is Part Of: Alzheimer's & dementia. Volume 16(2020)Supplement 5
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-12-07
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.044314
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: In recent years, many convolutional neural networks (CNN) have been proposed for the classification of Alzheimer's Disease (AD). Due to memory constraints, many of the proposed CNNs work at a 2D slice‐level or 3D patch‐level. Other subject‐level 3D CNNs which take a whole brain 3D MRI image as input require a long training time. Method: Here, we propose a lightweight subject‐level 3D CNN featuring dilated convolutions which allow the receptive field to be increased efficiently through a small number of layers. To comprehensively evaluate the generalizability of our proposed network, we performed four independent tests which includes testing on images from other ADNI individuals at various stages of the dementia, images acquired from other databases/sites (AIBL), images acquired using different protocols (OASIS) and longitudinal images acquired over a short period of time (MIRIAD). Result: We trained our network on the ADNI data, and we achieved a 5‐fold cross‐validated balanced accuracy of 88% in differentiating stable Dementia of the Alzheimer's type (sDAT) from stable normal controls (sNC). Our network showed 78.5% accuracy in classifying images of mild cognitive impairment (MCI) subjects acquired 2 years prior to conversion to DAT. We achieved an overall specificity of 79.5% and sensitivity 79.7% on the entire set of 7902 independent test images Conclusion: In this study, we constructed a lightweight 3D CNN network that converts the subject‐level image into a single AD dementia score to represent the disease progression. For estimating the generalization ability of the network to unseen data, independent testing is essential but is often lacking in studies using CNN for DAT classification. This makes it difficult to compare the performances achieved using different architectures. The result of our study highlights the competitive performance of our network and potential promise for generalization.