Disease Prediction based on Functional Connectomes using a Spatially-Informed Fused Lasso Support Vector Classifier

Substantial evidence indicates that major psychiatric disorders are associated with distributed neural dysconnectivity, leading to strong interest in using neuroimaging methods to accurately predict disorder status. In this work, we are specifically interested in a multivariate approach that uses features derived from whole-brain resting state functional connectomes. However, functional connectomes reside in a high dimensional space, which complicates model interpretation and introduces numerous statistical and computational challenges. Traditional feature selection techniques are used to reduce data dimensionality, but are blind to the spatial structure of the connectomes. Here, we propose a regularization framework where the $6$-D structure of the functional connectome (defined by pairs of points in $3$-D space) is explicitly taken into account via the sparse fused Lasso regularizer. Using this regularizer with the hinge-loss function leads to a structured sparse support vector classifier with embedded feature selection. We introduce a novel efficient optimization algorithm based on augmented Lagrangian and the classical alternating direction method, and demonstrate that the inner subproblems of the algorithm can be solved exactly and non-iteratively by coupling the variable splitting strategy with a data augmentation scheme. Experiments on simulated data and resting state scans from a large schizophrenia dataset show that our proposed approach can recover results that are more neuroscientifically informative than previous methods while preserving predictive power.