Sign-Consistency Based Variable Importance for Machine Learning in Brain Imaging

Vanessa Gómez-Verdejo,Jussi Tohka,Emilio Parrado-Hernández

doi:10.1007/s12021-019-9415-3

Vanessa Gómez-Verdejo, Jussi Tohka + Show 1 more

Open Access

https://doi.org/10.1007/s12021-019-9415-3

Copy DOI

Journal: Neuroinformatics	Publication Date: Mar 27, 2019
Citations: 16	License type: open-access

Affiliation: Carlos III University of Madrid

Abstract

An important problem that hinders the use of supervised classification algorithms for brain imaging is that the number of variables per single subject far exceeds the number of training subjects available. Deriving multivariate measures of variable importance becomes a challenge in such scenarios. This paper proposes a new measure of variable importance termed sign-consistency bagging (SCB). The SCB captures variable importance by analyzing the sign consistency of the corresponding weights in an ensemble of linear support vector machine (SVM) classifiers. Further, the SCB variable importances are enhanced by means of transductive conformal analysis. This extra step is important when the data can be assumed to be heterogeneous. Finally, the proposal of these SCB variable importance measures is completed with the derivation of a parametric hypothesis test of variable importance. The new importance measures were compared with a t-test based univariate and an SVM-based multivariate variable importances using anatomical and functional magnetic resonance imaging data. The obtained results demonstrated that the new SCB based importance measures were superior to the compared methods in terms of reproducibility and classification accuracy.

Highlights

Machine Learning (ML) is a powerful tool to characterize disease related alterations in brain structure and function.Alzheimer’s Disease Neuroimaging Initiative (ADNI) is a Group/ Institutional Author.Data used in preparation of this article were obtained from the Alzheimers Disease Neuroimaging Initiative (ADNI) database
To overcome the limitations of the regularized linear models and Random Forests2 (RFs), we introduce and study a new variable importance measure based on the sign consistency of the weights in an ensemble of linear Support Vector Machines (SVMs)
We have introduced and evaluated new variable importance measures, termed sign-consistency bagging (SCB) and SCBconf, based on sign consistency of classifier ensembles

Summary

Introduction

Machine Learning (ML) is a powerful tool to characterize disease related alterations in brain structure and function.Alzheimer’s Disease Neuroimaging Initiative (ADNI) is a Group/ Institutional Author.Data used in preparation of this article were obtained from the Alzheimers Disease Neuroimaging Initiative (ADNI) database (adni.loni.usc.edu). Given a training set of brain images and the associated class information, here a diagnosis of the subject, supervised ML algorithms learn a voxel-wise model that captures the class information from the brain images. This has direct applications to the design of imaging biomarkers, and the inferred models can be considered as multivariate, discriminative representations of the effect of the disease to brain images. This representation is fundamentally different from conventional brain maps that are constructed based on a voxel-by-voxel comparison of two groups of subjects (patients and controls) and the patterns of important voxels in these two types of analyses provide complementary information (Kerr et al 2014; Haufe et al 2014; Tohka et al 2016)

Methods

Results

Conclusion