Abstract

Brain image analyses have widely relied on univariate voxel-wise methods. In such analyses, brain images are first spatially registered to a common stereotaxic space, and then mass univariate statistical tests are performed in each voxel to detect significant group differences. However, the sensitivity of theses approaches is limited when the differences involve a combination of different brain structures. Recently, there has been a growing interest in support vector machines methods to overcome the limits of these analyses.

This talk will focus on machine learning methods for population analysis and patient classification in neuroimaging. First, we evaluated the performances of different classification strategies for the identification of patients with Alzheimer’s disease based on T1-weighted MRI . However, in these approaches, the specificity of neuroimaging data was not taken into account in the classification method per se. Brain images are indeed a prototypical case of structured data, whose structure is governed by the underlying anatomical and functional organization. Therefore we introduced a framework to introduce spatial and anatomical priors in SVM for brain image analysis based on regularization operators. The proposed framework was applied to the classification of brain magnetic resonance (MR) images (based on gray matter concentration maps and cortical thickness measures) from 137 patients with Alzheimer’s disease and 162 elderly controls. The results demonstrated that the proposed classifier generates less-noisy and consequently more interpretable feature maps with high classification performances.