High-Order Laplacian Regularized Low-Rank Representation for Multimodal Dementia Diagnosis.

Aimei Dong,Dinggang Shen,Zhigang Li,Mingxia Liu,Mingliang Wang

doi:10.3389/fnins.2021.634124

Abstract

Multimodal heterogeneous data, such as structural magnetic resonance imaging (MRI), positron emission tomography (PET), and cerebrospinal fluid (CSF), are effective in improving the performance of automated dementia diagnosis by providing complementary information on degenerated brain disorders, such as Alzheimer's prodromal stage, i.e., mild cognitive impairment. Effectively integrating multimodal data has remained a challenging problem, especially when these heterogeneous data are incomplete due to poor data quality and patient dropout. Besides, multimodal data usually contain noise information caused by different scanners or imaging protocols. The existing methods usually fail to well handle these heterogeneous and noisy multimodal data for automated brain dementia diagnosis. To this end, we propose a high-order Laplacian regularized low-rank representation method for dementia diagnosis using block-wise missing multimodal data. The proposed method was evaluated on 805 subjects (with incomplete MRI, PET, and CSF data) from the real Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort. Experimental results suggest the effectiveness of our method in three tasks of brain disease classification, compared with the state-of-the-art methods.

Highlights

Alzheimer’s disease (AD) is a highly prevalent and severe irreversible neurodegenerative disease and it has already devastated millions of lives in the world (Cuingnet et al, 2011)
Compared with several data completion/imputation methods (i.e., Zero, K-nearest neighbor (KNN), expectation maximization (EM), and singular value decomposition (SVD)), our algorithm has improved the accuracy results by about 3%
Compared with the other two deep learning methods, our algorithm improves the accuracy by about 3%, and improves the accuracy by 7% compared with the feature fusion method

Summary

Introduction

Alzheimer’s disease (AD) is a highly prevalent and severe irreversible neurodegenerative disease and it has already devastated millions of lives in the world (Cuingnet et al, 2011). AD is of an escalating epidemic and it is a tremendous challenge to global health care systems (Kuljis, 2010). It is estimated that the regular cost for caring for AD patients from families and health-care systems is up to $100 million every year (Reiman et al, 2010). It is estimated that the number of these patients nearly doubles every year and the number will be up to 115 million worldwide (Kuljis, 2010) and 13.8 million in the United States (Association et al, 2013) by 2050. Clinic and research show that the potential pathology of AD appears many years ahead of the onset of cognitive symptoms. Extensive studies pay attention to the automated diagnosis of AD and progression prediction of its prodrome, i.e., mild cognitive impairment (MCI), to delay the progress of AD

Methods

Results

Discussion

Conclusion