Abstract
The identification of abnormal cognitive decline at an early stage becomes an increasingly significant conundrum to physicians and is of major interest in the studies of mild cognitive impairment (MCI). Support vector machine (SVM) as a high-dimensional pattern classification technique is widely employed in neuroimaging research. However, the application of a single SVM classifier may be difficult to achieve the excellent classification performance because of the small-sample size and noise of imaging data. To address this issue, we propose a novel method of the weighted random support vector machine cluster (WRSVMC) in which multiple SVMs were built and different weights were given to corresponding SVMs with different classification performances. We evaluated our algorithm on resting state functional magnetic resonance imaging (RS-fMRI) data of 93 MCI patients and 105 healthy controls (HC) from the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort. The maximum accuracy given by the WRSVMC is 87.67%, demonstrating excellent diagnostic power. Furthermore, the most discriminative brain areas have been found out as follows: gyrus rectus (REC.L), precentral gyrus (PreCG.R), olfactory cortex (OLF.L), and middle occipital gyrus (MOG.R). These findings of the paper provide a new perspective for the clinical diagnosis of MCI.
Highlights
Mild cognitive impairment (MCI) is a clinical entity which represents a state of slightly cognitive deficits for age and education, but does not markedly affect activities of daily life [1, 2]
Compared to a single Support vector machine (SVM) classifier, the weighted random support vector machine cluster (WRSVMC) has the following advantages: [1] The WRSVMC is robust because it consists of a great deal of SVM classifiers; [2] The classification accuracy of the WRSVMC is improved because the influences of strong SVM base classifiers are enhanced by a weighted method; [3] The abnormal brain areas could be found out using the WRSVMC based on the optimal subset of features; [4] The WRSVMC achieves an high accuracy of 87.67%, indicating that the abnormal brain areas which we have found were considerably convincing
The two-sample t-test is conducted to examine the differences of the WRSVMC/random SVM cluster (RSVMC) and WRSVMC/random forest (RF) respectively and the P-values are close to 0.00 and 0.00, which indicates that the differences between the WRSVMC and other two classification methods are statistical significance
Summary
Mild cognitive impairment (MCI) is a clinical entity which represents a state of slightly cognitive deficits for age and education, but does not markedly affect activities of daily life [1, 2]. The rate of MCI patients who progress to AD is between 10 and 15% per year [4], implying that MCI may be a high-risk state for developing AD dementia. It is crucial to identify MCI patients and explore pathological changes in their brains, in order to offer timely treatment and slow down the transition from MCI to AD. Neuroimaging techniques play increasingly important roles in the investigation of brain dysfunctions of MCI patients [6]. The resting-state functional magnetic resonance imaging (RS-fMRI) may be one of the most popular brain imaging techniques due to its numerous
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
