AUTHOR=Ma Peixian , Wang Jing , Zhou Zhiguo , Chen C. L. Philip ,  The Alzheimer's Disease Neuroimaging Initiative , Duan Junwei 

TITLE=Development and validation of a deep-broad ensemble model for early detection of Alzheimer's disease

JOURNAL=Frontiers in Neuroscience

VOLUME=Volume 17 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2023.1137557

DOI=10.3389/fnins.2023.1137557

ISSN=1662-453X

ABSTRACT=Deep learning methods are widely applied in early detection of Alzheimer's disease. However, training a suitable deep model for this task has a certain risk of overfitting, requires expensive hardware resources and consumes a lot of time. These problems restrict the further use of artificial intelligence in Alzheimer's disease diagnosis.We constructed a deep-broad ensemble model based on 3D residual convolution module and Broad Learning System(BLS) for early detection of Alzheimer's disease. We utilized the publicly available MRI image dataset from Alzheimer's Disease Neuroimaging Initiative (ADNI) to classify patients into three different states. We demonstrate that the proposed deep-broad ensemble model perform better than previous deep models, including ResNet and VoxCNN in terms of accuracy, sensitivity and F1-scores.In addition, we demonstrate that the proposed model does not require the pre-training process of the traditional deep model, which greatly reduces the training time and hardware dependency.