Early identification of mild cognitive impairment using incomplete random forest-robust support vector machine and FDG-PET imaging

• Published in: Computerized medical imaging and graphics Vol. 60; pp. 35 - 41
• Main Authors: Lu, Shen, Xia, Yong, Cai, Weidong, Fulham, Michael, Feng, David Dagan
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.09.2017; Elsevier Science Ltd
• Subjects: Aged; Alzheimer Disease - diagnostic imaging; Alzheimer's disease; Brain; Cognition & reasoning; Cognitive ability; Cognitive Dysfunction - diagnostic imaging; Cortex; Dementia disorders; FDG-PET; Feature extraction; Female; Fluorodeoxyglucose F18; Humans; Identification; Impairment; Internal Medicine; Male; Medical imaging; Mild cognitive impairment; Neurodegenerative diseases; Optimization; Other; Population (statistical); Positron emission; Positron emission tomography; Prodromal Symptoms; Robust optimization; Robustness; Support Vector Machine; Tomography; Uncertainty; FDG-PET; Robust optimization; Alzheimer’s disease; Mild cognitive impairment
• ISSN: 0895-6111; 1879-0771; 1879-0771
• DOI: 10.1016/j.compmedimag.2017.01.001

•A dementia classification method using robust optimization and FDG-PET imaging is proposed.•Three different FDG-PET image datasets were used for evaluation.•The proposed method outperforms several state-of-the-art transfer learning schemes in our comparative experiments.•The performance of our method is generally not sensitive to the variation of regularization parameters when a non-linear kernel is used. Alzheimer’s disease (AD) is the most common type of dementia and will be an increasing health problem in society as the population ages. Mild cognitive impairment (MCI) is considered to be a prodromal stage of AD. The ability to identify subjects with MCI will be increasingly important as disease modifying therapies for AD are developed. We propose a semi-supervised learning method based on robust optimization for the identification of MCI from [18F]Fluorodeoxyglucose PET scans. We extracted three groups of spatial features from the cortical and subcortical regions of each FDG-PET image volume. We measured the statistical uncertainty related to these spatial features via transformation using an incomplete random forest and formulated the MCI identification problem under a robust optimization framework. We compared our approach to other state-of-the-art methods in different learning schemas. Our method outperformed the other techniques in the ability to separate MCI from normal controls.