King's College London

Research portal

Secure multivariate large-scale multi-centric analysis through on-line learning: An imaging genetics case study

Research output: Chapter in Book/Report/Conference proceedingConference paper

Standard

Secure multivariate large-scale multi-centric analysis through on-line learning : An imaging genetics case study. / Lorenzi, Marco; Gutman, Boris; Thompson, Paul M.; Alexander, Daniel C.; Ourselin, Sebastien; Altmann, Andre.

12th International Symposium on Medical Information Processing and Analysis. Vol. 10160 SPIE, 2017. 1016016.

Research output: Chapter in Book/Report/Conference proceedingConference paper

Harvard

Lorenzi, M, Gutman, B, Thompson, PM, Alexander, DC, Ourselin, S & Altmann, A 2017, Secure multivariate large-scale multi-centric analysis through on-line learning: An imaging genetics case study. in 12th International Symposium on Medical Information Processing and Analysis. vol. 10160, 1016016, SPIE, 12th International Symposium on Medical Information Processing and Analysis, SIPAIM 2016, Tandil, Argentina, 5/12/2016. https://doi.org/10.1117/12.2256799

APA

Lorenzi, M., Gutman, B., Thompson, P. M., Alexander, D. C., Ourselin, S., & Altmann, A. (2017). Secure multivariate large-scale multi-centric analysis through on-line learning: An imaging genetics case study. In 12th International Symposium on Medical Information Processing and Analysis (Vol. 10160). [1016016] SPIE. https://doi.org/10.1117/12.2256799

Vancouver

Lorenzi M, Gutman B, Thompson PM, Alexander DC, Ourselin S, Altmann A. Secure multivariate large-scale multi-centric analysis through on-line learning: An imaging genetics case study. In 12th International Symposium on Medical Information Processing and Analysis. Vol. 10160. SPIE. 2017. 1016016 https://doi.org/10.1117/12.2256799

Author

Lorenzi, Marco ; Gutman, Boris ; Thompson, Paul M. ; Alexander, Daniel C. ; Ourselin, Sebastien ; Altmann, Andre. / Secure multivariate large-scale multi-centric analysis through on-line learning : An imaging genetics case study. 12th International Symposium on Medical Information Processing and Analysis. Vol. 10160 SPIE, 2017.

Bibtex Download

@inbook{5e983c290230417599576c6932cb8fea,
title = "Secure multivariate large-scale multi-centric analysis through on-line learning: An imaging genetics case study",
abstract = "State-of-the-art data analysis methods in genetics and related fields have advanced beyond massively univariate analyses. However, these methods suffer from the limited amount of data available at a single research site. Recent large-scale multi-centric imaging-genetic studies, such as ENIGMA, have to rely on meta-analysis of mass univariate models to achieve critical sample sizes for uncovering statistically significant associations. Indeed, model parameters, but not data, can be securely and anonymously shared between partners. We propose here partial least squares (PLS) as a multivariate imaging-genetics model in meta-studies. In particular, we propose an online estimation approach to partial least squares for the sequential estimation of the model parameters in data batches, based on an approximation of the singular value decomposition (SVD) of partitioned covariance matrices. We applied the proposed approach to the challenging problem of modeling the association between 1,167,117 genetic markers (SNPs, single nucleotide polymorphisms) and the brain cortical and sub-cortical atrophy (354,804 anatomical surface features) in a cohort of 639 individuals from the Alzheimer's Disease Neuroimaging Initiative. We compared two different modeling strategies (sequential-and meta-PLS) to the classic non-distributed PLS. Both strategies exhibited only minimal approximation errors of model parameters. The proposed approaches pave the way to the application of multivariate models in large scale imaging-genetics meta-studies, and may lead to novel understandings of the complex brain phenotype-genotype interactions.",
keywords = "Imaging-genetics, Meta analysis, Online learning, Partial least squares",
author = "Marco Lorenzi and Boris Gutman and Thompson, {Paul M.} and Alexander, {Daniel C.} and Sebastien Ourselin and Andre Altmann",
year = "2017",
month = "1",
day = "1",
doi = "10.1117/12.2256799",
language = "English",
volume = "10160",
booktitle = "12th International Symposium on Medical Information Processing and Analysis",
publisher = "SPIE",

}

RIS (suitable for import to EndNote) Download

TY - CHAP

T1 - Secure multivariate large-scale multi-centric analysis through on-line learning

T2 - An imaging genetics case study

AU - Lorenzi, Marco

AU - Gutman, Boris

AU - Thompson, Paul M.

AU - Alexander, Daniel C.

AU - Ourselin, Sebastien

AU - Altmann, Andre

PY - 2017/1/1

Y1 - 2017/1/1

N2 - State-of-the-art data analysis methods in genetics and related fields have advanced beyond massively univariate analyses. However, these methods suffer from the limited amount of data available at a single research site. Recent large-scale multi-centric imaging-genetic studies, such as ENIGMA, have to rely on meta-analysis of mass univariate models to achieve critical sample sizes for uncovering statistically significant associations. Indeed, model parameters, but not data, can be securely and anonymously shared between partners. We propose here partial least squares (PLS) as a multivariate imaging-genetics model in meta-studies. In particular, we propose an online estimation approach to partial least squares for the sequential estimation of the model parameters in data batches, based on an approximation of the singular value decomposition (SVD) of partitioned covariance matrices. We applied the proposed approach to the challenging problem of modeling the association between 1,167,117 genetic markers (SNPs, single nucleotide polymorphisms) and the brain cortical and sub-cortical atrophy (354,804 anatomical surface features) in a cohort of 639 individuals from the Alzheimer's Disease Neuroimaging Initiative. We compared two different modeling strategies (sequential-and meta-PLS) to the classic non-distributed PLS. Both strategies exhibited only minimal approximation errors of model parameters. The proposed approaches pave the way to the application of multivariate models in large scale imaging-genetics meta-studies, and may lead to novel understandings of the complex brain phenotype-genotype interactions.

AB - State-of-the-art data analysis methods in genetics and related fields have advanced beyond massively univariate analyses. However, these methods suffer from the limited amount of data available at a single research site. Recent large-scale multi-centric imaging-genetic studies, such as ENIGMA, have to rely on meta-analysis of mass univariate models to achieve critical sample sizes for uncovering statistically significant associations. Indeed, model parameters, but not data, can be securely and anonymously shared between partners. We propose here partial least squares (PLS) as a multivariate imaging-genetics model in meta-studies. In particular, we propose an online estimation approach to partial least squares for the sequential estimation of the model parameters in data batches, based on an approximation of the singular value decomposition (SVD) of partitioned covariance matrices. We applied the proposed approach to the challenging problem of modeling the association between 1,167,117 genetic markers (SNPs, single nucleotide polymorphisms) and the brain cortical and sub-cortical atrophy (354,804 anatomical surface features) in a cohort of 639 individuals from the Alzheimer's Disease Neuroimaging Initiative. We compared two different modeling strategies (sequential-and meta-PLS) to the classic non-distributed PLS. Both strategies exhibited only minimal approximation errors of model parameters. The proposed approaches pave the way to the application of multivariate models in large scale imaging-genetics meta-studies, and may lead to novel understandings of the complex brain phenotype-genotype interactions.

KW - Imaging-genetics

KW - Meta analysis

KW - Online learning

KW - Partial least squares

UR - http://www.scopus.com/inward/record.url?scp=85014678233&partnerID=8YFLogxK

U2 - 10.1117/12.2256799

DO - 10.1117/12.2256799

M3 - Conference paper

AN - SCOPUS:85014678233

VL - 10160

BT - 12th International Symposium on Medical Information Processing and Analysis

PB - SPIE

ER -

View graph of relations

© 2018 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454