Support Vector Machine Classification of Spontaneous Cognition Using Whole-Brain Resting-State Functional Connectivity

Ying-Hui Chou¹, Pooja Gaur², Carol P. Weingarten¹, Mei-Lan Chu¹, David Madden¹, Allen W. Song¹, Nan-Kuei Chen³

In this study, we demonstrated that the behavior-based connectivity analysis and support vector machine methods can be used to decode the whole-brain resting-state functional connectivity patterns and classify individuals who reported inner language as the dominant mental activity during resting-state fMRI scan from those who did not with a sensitivity of 0.88 and a specificity of 0.9. Our findings can lead to a better understanding of variations in resting-state fMRI signals and their dependence on the spontaneous cognition/mind wandering.