Two-way regularization for MEG source reconstruction via multilevel coordinate descent

Handle URI:
http://hdl.handle.net/10754/600114
Title:
Two-way regularization for MEG source reconstruction via multilevel coordinate descent
Authors:
Siva Tian, Tian; Huang, Jianhua Z.; Shen, Haipeng
Abstract:
Magnetoencephalography (MEG) source reconstruction refers to the inverse problem of recovering the neural activity from the MEG time course measurements. A spatiotemporal two-way regularization (TWR) method was recently proposed by Tian et al. to solve this inverse problem and was shown to outperform several one-way regularization methods and spatiotemporal methods. This TWR method is a two-stage procedure that first obtains a raw estimate of the source signals and then refines the raw estimate to ensure spatial focality and temporal smoothness using spatiotemporal regularized matrix decomposition. Although proven to be effective, the performance of two-stage TWR depends on the quality of the raw estimate. In this paper we directly solve the MEG source reconstruction problem using a multivariate penalized regression where the number of variables is much larger than the number of cases. A special feature of this regression is that the regression coefficient matrix has a spatiotemporal two-way structure that naturally invites a two-way penalty. Making use of this structure, we develop a computationally efficient multilevel coordinate descent algorithm to implement the method. This new one-stage TWR method has shown its superiority to the two-stage TWR method in three simulation studies with different levels of complexity and a real-world MEG data analysis. © 2013 Wiley Periodicals, Inc., A Wiley Company.
Citation:
Siva Tian T, Huang JZ, Shen H (2013) Two-way regularization for MEG source reconstruction via multilevel coordinate descent. Statistical Analy Data Mining 6: 545–556. Available: http://dx.doi.org/10.1002/sam.11210.
Publisher:
Wiley-Blackwell
Journal:
Statistical Analysis and Data Mining
KAUST Grant Number:
KUS-CI-016-04
Issue Date:
Dec-2013
DOI:
10.1002/sam.11210
Type:
Article
ISSN:
1932-1864
Sponsors:
This work was supported in part by NCI (CA57030), NSF (DMS-09-07170, DMS-10-07618, CMMI-0800575, DMS-11-06912, DMS-12-08952, and DMS-12-08786), NIDA (1 RC1 DA029425-01), and King Abdullah University of Science and Technology (KUS-CI-016-04).
Appears in Collections:
Publications Acknowledging KAUST Support

Full metadata record

DC FieldValue Language
dc.contributor.authorSiva Tian, Tianen
dc.contributor.authorHuang, Jianhua Z.en
dc.contributor.authorShen, Haipengen
dc.date.accessioned2016-02-28T06:42:57Zen
dc.date.available2016-02-28T06:42:57Zen
dc.date.issued2013-12en
dc.identifier.citationSiva Tian T, Huang JZ, Shen H (2013) Two-way regularization for MEG source reconstruction via multilevel coordinate descent. Statistical Analy Data Mining 6: 545–556. Available: http://dx.doi.org/10.1002/sam.11210.en
dc.identifier.issn1932-1864en
dc.identifier.doi10.1002/sam.11210en
dc.identifier.urihttp://hdl.handle.net/10754/600114en
dc.description.abstractMagnetoencephalography (MEG) source reconstruction refers to the inverse problem of recovering the neural activity from the MEG time course measurements. A spatiotemporal two-way regularization (TWR) method was recently proposed by Tian et al. to solve this inverse problem and was shown to outperform several one-way regularization methods and spatiotemporal methods. This TWR method is a two-stage procedure that first obtains a raw estimate of the source signals and then refines the raw estimate to ensure spatial focality and temporal smoothness using spatiotemporal regularized matrix decomposition. Although proven to be effective, the performance of two-stage TWR depends on the quality of the raw estimate. In this paper we directly solve the MEG source reconstruction problem using a multivariate penalized regression where the number of variables is much larger than the number of cases. A special feature of this regression is that the regression coefficient matrix has a spatiotemporal two-way structure that naturally invites a two-way penalty. Making use of this structure, we develop a computationally efficient multilevel coordinate descent algorithm to implement the method. This new one-stage TWR method has shown its superiority to the two-stage TWR method in three simulation studies with different levels of complexity and a real-world MEG data analysis. © 2013 Wiley Periodicals, Inc., A Wiley Company.en
dc.description.sponsorshipThis work was supported in part by NCI (CA57030), NSF (DMS-09-07170, DMS-10-07618, CMMI-0800575, DMS-11-06912, DMS-12-08952, and DMS-12-08786), NIDA (1 RC1 DA029425-01), and King Abdullah University of Science and Technology (KUS-CI-016-04).en
dc.publisherWiley-Blackwellen
dc.subjectCoordinate descenten
dc.subjectInverse problemen
dc.subjectSpatiotemporalen
dc.subjectTwo-way regularizationen
dc.titleTwo-way regularization for MEG source reconstruction via multilevel coordinate descenten
dc.typeArticleen
dc.identifier.journalStatistical Analysis and Data Miningen
dc.contributor.institutionUniversity of Houston, Houston, United Statesen
dc.contributor.institutionTexas A and M University, College Station, United Statesen
dc.contributor.institutionThe University of North Carolina at Chapel Hill, Chapel Hill, United Statesen
kaust.grant.numberKUS-CI-016-04en
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.