IRUS Total

p-Value combiners for graphical modelling of EEG data in thefrequency domain

File Description SizeFormat 
SchneiderLuftman.pdfPublished version1.46 MBAdobe PDFView/Open
Title: p-Value combiners for graphical modelling of EEG data in thefrequency domain
Authors: Schneider-Luftman, D
Item Type: Journal Article
Abstract: Background: In the graphical modelling of brain data, we are interested in estimating connectivitybetween various regions of interest, and evaluating statistical significance in order to derive a networkmodel. This process involves aggregating results across frequency ranges and several patients, in orderto obtain an overall result that can serve to construct a graph.New method: In this paper, we propose a method based on p-value combiners, which have never beenused in applications to EEG data analysis. This new method is split into two aspects: frequency-wide testsand group-wide tests. The first step can be effectively adjusted to control for false detection rate.Results: This two-step protocol is applied to EEG data collected from distinct groups of mental healthpatients, in order to draw graphical models for each group and highlight structural connectivity differ-ences. Using the method proposed, we show that it is possible to reliably achieve this while effectivelycontrolling for false connections detection.Comparison with existing method(s): Conventionally, the Holm’s Stepdown procedure is used for this typeof problem, as it is robust to type I errors. However, it is known to be conservative and prone to falsenegatives. Furthermore, unlike the proposed methods, it does not directly output a decision rule onwhether to accept or reject a statement.Conclusions: The proposed methodology offers significant improvements over the stepdown procedurein terms of error rate and false negative rate across the network models, as well as in term of applicability.
Issue Date: 21-Jul-2016
Date of Acceptance: 18-Jul-2016
URI: http://hdl.handle.net/10044/1/38847
DOI: https://dx.doi.org/10.1016/j.jneumeth.2016.07.006
ISSN: 1872-678X
Publisher: Elsevier
Start Page: 92
End Page: 106
Journal / Book Title: Journal of Neuroscience Methods
Volume: 271
Copyright Statement: © 2016 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Keywords: 1109 Neurosciences
1702 Cognitive Science
Publication Status: Published
Appears in Collections:School of Public Health