TY - JOUR
T1 - Addressing Confounding in Predictive Models with an Application to Neuroimaging
AU - Linn, Kristin A.
AU - Gaonkar, Bilwaj
AU - Doshi, Jimit
AU - Davatzikos, Christos
AU - Shinohara, Russell T.
N1 - Funding Information:
The authors would like to acknowledge funding by NIH grant R01NS085211 and a seed grant from the Center for Biomedical Image Computing and Analytics at the University of Pennsylvania. This work represents the opinions of the researchers and not necessarily that of the granting institutions.
Publisher Copyright:
© 2016 by De Gruyter.
PY - 2016/5/1
Y1 - 2016/5/1
N2 - Understanding structural changes in the brain that are caused by a particular disease is a major goal of neuroimaging research. Multivariate pattern analysis (MVPA) comprises a collection of tools that can be used to understand complex disease efxcfects across the brain. We discuss several important issues that must be considered when analyzing data from neuroimaging studies using MVPA. In particular, we focus on the consequences of confounding by non-imaging variables such as age and sex on the results of MVPA. After reviewing current practice to address confounding in neuroimaging studies, we propose an alternative approach based on inverse probability weighting. Although the proposed method is motivated by neuroimaging applications, it is broadly applicable to many problems in machine learning and predictive modeling. We demonstrate the advantages of our approach on simulated and real data examples.
AB - Understanding structural changes in the brain that are caused by a particular disease is a major goal of neuroimaging research. Multivariate pattern analysis (MVPA) comprises a collection of tools that can be used to understand complex disease efxcfects across the brain. We discuss several important issues that must be considered when analyzing data from neuroimaging studies using MVPA. In particular, we focus on the consequences of confounding by non-imaging variables such as age and sex on the results of MVPA. After reviewing current practice to address confounding in neuroimaging studies, we propose an alternative approach based on inverse probability weighting. Although the proposed method is motivated by neuroimaging applications, it is broadly applicable to many problems in machine learning and predictive modeling. We demonstrate the advantages of our approach on simulated and real data examples.
KW - Multivariate pattern analysis (MVPA)
KW - confounding
KW - inverse probability weighting
KW - machine learning
KW - predictive modeling
KW - structural magnetic resonance imaging (MRI)
KW - support vector machine (SVM)
UR - http://www.scopus.com/inward/record.url?scp=84975299432&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84975299432&partnerID=8YFLogxK
U2 - 10.1515/ijb-2015-0030
DO - 10.1515/ijb-2015-0030
M3 - Article
C2 - 26641972
AN - SCOPUS:84975299432
SN - 1557-4679
VL - 12
SP - 31
EP - 44
JO - International Journal of Biostatistics
JF - International Journal of Biostatistics
IS - 1
ER -