Independent Component Analysis-Based Classification of Alzheimer's Disease MRI Data

Article type: Research Article

Authors: Yang, Wenlu^{a; f} | Lui, Ronald L.M.^b | Gao, Jia-Hong^c | Chan, Tony F.^d | Yau, Shing-Tung^b | Sperling, Reisa A.^e | Huang, Xudong^{f; g; *}

Affiliations: [a] Department of Electrical Engineering, Information Engineering College, Shanghai Maritime University, Shanghai, China | [b] Department of Mathematics, Harvard University, Cambridge, MA, USA | [c] Brain Research Imaging Center, The University of Chicago, Chicago, IL, USA | [d] The Hong Kong University of Science and Technology, Hong Kong, China | [e] Memory Disorders Unit, Department of Neurology, Brigham and Women's Hospital and Harvard Medical School, Boston, MA, USA | [f] Biomedical Informatics and Cheminformatics Group, Conjugate and Medical Chemistry Laboratory, Division of Nuclear Medicine and Molecular Imaging and Center for Advanced Medical Imaging, Department of Radiology, Brigham and Women's Hospital and Harvard Medical School, Boston, MA, USA | [g] Neurochemistry Laboratory, Department of Psychiatry, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA

Correspondence: [*] Correspondence to: Xudong Huang, Ph.D., Department of Radiology, Brigham and Women's Hospital and Harvard Medical School, 75 Francis Street, Boston, MA 02115, USA. Tel: +1 617 582 4711; Fax: +1 617 582 0004; E-mail: [email protected].

Abstract: There is an unmet medical need to identify neuroimaging biomarkers that allow us to accurately diagnose and monitor Alzheimer's disease (AD) at its very early stages and to assess the response to AD-modifying therapies. To a certain extent, volumetric and functional magnetic resonance imaging (fMRI) studies can detect changes in structure, cerebral blood flow, and blood oxygenation that distinguish AD and mild cognitive impairment (MCI) subjects from healthy control (HC) subjects. However, it has been challenging to use fully automated MRI analytic methods to identify potential AD neuroimaging biomarkers. We have thus proposed a method based on independent component analysis (ICA) for studying potential AD-related MR image features that can be coupled with the use of support vector machine (SVM) for classifying scans into categories of AD, MCI, and HC subjects. The MRI data were selected from the Open Access Series of Imaging Studies (OASIS) and the Alzheimer's Disease Neuroimaging Initiative databases. The experimental results showed that the ICA method coupled with SVM classifier can differentiate AD and MCI patients from HC subjects, although further methodological improvement in the analytic method and inclusion of additional variables may be required for optimal classification.

Keywords: Alzheimer's disease, independent component analysis, magnetic resonance imaging, mild cognitive impairment, neuroimaging biomarker, support vector machine

DOI: 10.3233/JAD-2011-101371

Journal: Journal of Alzheimer's Disease, vol. 24, no. 4, pp. 775-783, 2011