Multi-stream multi-scale deep convolutional networks for Alzheimer’s disease detection using MR images

• Published in: Neurocomputing (Amsterdam) Vol. 350; pp. 60 - 69
• Main Authors: Ge, Chenjie, Qu, Qixun, Gu, Irene Yu-Hua, Jakola, Asgeir Store
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 20.07.2019
• Subjects: Alzheimer's disease detection; and Signal Processing; classification; Computer Science; Deep convolutional networks; Deep learning; Feature boosting and dimension reduction; Feature fusion; MR images; Multi-scale; Multi-scale feature learning; Neurosciences; Neurovetenskaper; OCESSING (ICASSP)IEEE International Conference on Acoustics; Speech; Tissue region; Deep learning; MR images; Deep convolutional networks; Feature fusion; Tissue region; Multi-scale feature learning; Feature boosting and dimension reduction; Alzheimer’s disease detection
• ISSN: 0925-2312; 1872-8286; 1872-8286
• DOI: 10.1016/j.neucom.2019.04.023

This paper addresses the issue of Alzheimer’s disease (AD) detection from Magnetic Resonance Images (MRIs). Existing AD detection methods rely on global feature learning from the whole brain scans, while depending on the tissue types, AD related features in different tissue regions, e.g. grey matter (GM), white matter (WM), and cerebrospinal fluid (CSF), show different characteristics. In this paper, we propose a deep learning method for multi-scale feature learning based on segmented tissue areas. A novel deep 3D multi-scale convolutional network scheme is proposed to generate multi-resolution features for AD detection. The proposed scheme employs several parallel 3D multi-scale convolutional networks, each applying to individual tissue regions (GM, WM and CSF) followed by feature fusions. The proposed fusion is applied in two separate levels: the first level fusion is applied on different scales within the same tissue region, and the second level is on different tissue regions. To further reduce the dimensions of features and mitigate overfitting, a feature boosting and dimension reduction method, XGBoost, is utilized before the classification. The proposed deep learning scheme has been tested on a moderate open dataset of ADNI (1198 scans from 337 subjects), with excellent test performance on randomly partitioned datasets (best 99.67%, average 98.29%), and good test performance on subject-separated partitioned datasets (best 94.74%, average 89.51%). Comparisons with state-of-the-art methods are also included.