Sparse and shrunken estimates of MRI networks in the brain and their influence on network properties

Rafael Romero-Garcia, Line Katrine Harder Clemmensen

Research output: Contribution to journalConference articleResearchpeer-review

Abstract

Estimation of morphometric relationships between cortical regions is a widely used approach to identify and characterize structural connectivity. The elevated number of regions that can be considered in a whole-brain correlation analysis might lead to overfitted models. However, the overfitting can be avoided by using regularization methods. We found that, as expected, non-regularized correlations had low variability when a scarce number of variables were considered.
However, a slight increase of variables led to an increase of variance of several magnitude orders. On the other hand, the regularized approaches showed more stable results with a relative low variance at the expense of a little bias. Interestingly, topological properties as local and global efficiency estimated in networks constructed from traditional non-regularized correlations also showed higher variability when compared to those from regularized networks. Our findings suggest that a population-based connectivity study can achieve a more robust description of cortical topology through regularization of the correlation estimates. Four regularization methods were examined: Two with shrinkage (Ridge and Schäfer’s shrinkage), one with sparsity (Lasso) and one with both shrinkage and sparsity (Elastic net). Furthermore, the different regularizations resulted in different correlation estimates as well as network properties. The shrunken estimates resulted in lower variance of the estimates than the sparse estimates.
Original languageEnglish
Article number903428
JournalProceedings of SPIE, the International Society for Optical Engineering
Volume9034
Number of pages7
ISSN1605-7422
DOIs
Publication statusPublished - 2014
EventSPIE Medical Imaging 2014 - Town & Country Resort and Convention Center, San Diego, California, United States
Duration: 15 Feb 201420 Feb 2014
http://spie.org/x12166.xml

Conference

ConferenceSPIE Medical Imaging 2014
LocationTown & Country Resort and Convention Center
CountryUnited States
CitySan Diego, California
Period15/02/201420/02/2014
Internet address

Keywords

  • Cortical network
  • Network properties
  • MRI
  • Partial correlation coefficients
  • Regularization
  • Shrinkage estimators
  • Sparse estimators
  • Structural connectivity

Fingerprint Dive into the research topics of 'Sparse and shrunken estimates of MRI networks in the brain and their influence on network properties'. Together they form a unique fingerprint.

Cite this