Assessment of Alzheimer-related Pathologies of Dementia Using Machine Learning Feature Selection

Abstract

AbstractAlthough a variety of brain lesions may contribute to the pathological diagnosis of dementia, the relationship of these lesions to dementia, how they interact and how to quantify them remain uncertain. Systematically assessing neuropathological measures in relation to the cognitive and functional definitions of dementia may enable the development of better diagnostic systems and treatment targets. The objective of this study is to apply machine learning approaches for feature selection to identify key features of Alzheimer-related pathologies associated with dementia. We applied machine learning techniques for feature ranking and classification as an unbiased comparison of neuropathological features and assessment of their diagnostic performance using a cohort (n=186) from the Cognitive Function and Ageing Study (CFAS). Seven feature ranking methods using different information criteria consistently ranked 22 out of the 34 neuropathology features for importance to dementia classification. Braak neurofibrillary tangle stage, Beta-amyloid and cerebral amyloid angiopathy features were the most highly ranked, although were highly correlated with each other. The best performing dementia classifier using the top eight ranked neuropathology features achieved 79% sensitivity, 69% specificity, and 75% precision. A substantial proportion (40.4%) of dementia cases was consistently misclassified by all seven algorithms and any combination of the 22 ranked features. These results highlight the potential of using machine learning to identify key indices of plaque, tangle and cerebral amyloid angiopathy burdens that may be useful for the classification of dementia.