Brain pathway-based discriminatory analysis for alzheimer’s disease

Abstract

The dysfunction of neural activity in Alzheimer`s disease could cause disconnection at both the structural and functional system of the brain. The advent of resting-state (RS) functional magnetic resonance imaging (fMRI) technology has made it possible to classify Alzheimer`s disease states based on the quantitative activity indices of brain regions. While the functional connectivity between brain regions has been regarded as a more effective index than sole activity levels of individual brain regions, current connectivity-based classification techniques suffer from limited reproducibility due to the need for prior knowledge on discriminative brain regions and intrinsic heterogeneity in the course of AD progression. Actually, similar challenges have been already addressed in molecular bioinformatics communities. They have achieved higher and reproducible classification accuracy and have identified interpretable markers by incorporating molecular pathway information in their classification. We have adopted a similar strategy to the RS-fMRI-based AD classification problem. After collecting various functional brain pathways from literature, we have quantified which pathways show significantly different activity levels between AD patients and healthy subjects. In this work, we used the brain pathway information covering the intrinsic brain function, such as cognitive, motor, and sensory function. Then, the brain pathway activity scores were acquired by greed search algorithm. We have shown that three different classification models, Naive Bayes (NB), linear logistic regression, and support vector machine (SVM) using such pathway activity scores have achieved robust classification performance. Moreover, discriminatory pathways between AD patients and healthy subjects may facilitate the interpretation of functional alterations in the course of AD progression.