Locally Linear Embedding of Anatomical Connectivity for Classification

Abstract

In this chapter, we aimed at investigating the alterations of the whole-brain anatomical connectivity in major depression with a machine learning approach combined local linear embedding with support vector machines. Brain anatomical networks were extracted from diffusion magnetic resonance images obtained from 22 first-episode, treatment-naive adults with major depressive disorder and 26 matched healthy controls. Using the machine learning approach, we differentiated depressed patients from healthy controls based on their whole-brain anatomical connectivity patterns and identified most discriminating features representing the between-group differences. Classification results showed that 91.7% (patients = 86.4%, controls = 96.2%; permutation test, p < 0.0001) of subjects were correctly classified via leave-one-out cross-validation. Moreover, the strengths of all the most discriminating connections were increased in depressed patients relative to the controls, and these connections were primarily located within the cortical-limbic network, especially the frontal-limbic network. These results not only provide initial steps toward the development of neurobiological diagnostic markers for major depressive disorder but also suggest that abnormal cortical-limbic anatomical network may contribute to the anatomical basis of emotion dysregulation and cognitive impairments of this disease.