Bayesian multi-dipole modelling in the frequency domain

Gianvittorio Luria, Dunja Duran, Elisa Visani, Sara Sommariva, Fabio Rotondi, Davide Rossi Sebastiano, Ferruccio Panzica, Michele Piana, Alberto Sorrentino

Research output: Contribution to journalArticle

Abstract

Background: Magneto- and Electro-encephalography record the electromagnetic field generated by neural currents with high temporal frequency and good spatial resolution, and are therefore well suited for source localization in the time and in the frequency domain. In particular, localization of the generators of neural oscillations is very important in the study of cognitive processes in the healthy and in the pathological brain. New method: We introduce the use of a Bayesian multi-dipole localization method in the frequency domain. Given the Fourier Transform of the data at one or multiple frequencies and/or trials, the algorithm approximates numerically the posterior distribution with Monte Carlo techniques. Results: We use synthetic data to show that the proposed method behaves well under a wide range of experimental conditions, including low signal-to-noise ratios and correlated sources. We use dipole clusters to mimic the effect of extended sources. In addition, we test the algorithm on real MEG data to confirm its feasibility. Comparison with existing method(s): Throughout the whole study, DICS (Dynamic Imaging of Coherent Sources) is used systematically as a benchmark. The two methods provide similar general pictures; the posterior distributions of the Bayesian approach contain much richer information at the price of a higher computational cost. Conclusions: The Bayesian method described in this paper represents a reliable approach for localization of multiple dipoles in the frequency domain.

Original languageEnglish
Pages (from-to)27-36
Number of pages10
JournalJournal of Neuroscience Methods
Volume312
DOIs
Publication statusPublished - Jan 15 2019

Fingerprint

Bayes Theorem
Benchmarking
Electromagnetic Fields
Signal-To-Noise Ratio
Fourier Analysis
Costs and Cost Analysis
Brain

Keywords

  • Bayesian methods
  • EEG/MEG
  • Oscillatory brain activity
  • Sequential Monte Carlo
  • Source modelling

ASJC Scopus subject areas

  • Neuroscience(all)

Cite this

Bayesian multi-dipole modelling in the frequency domain. / Luria, Gianvittorio; Duran, Dunja; Visani, Elisa; Sommariva, Sara; Rotondi, Fabio; Rossi Sebastiano, Davide; Panzica, Ferruccio; Piana, Michele; Sorrentino, Alberto.

In: Journal of Neuroscience Methods, Vol. 312, 15.01.2019, p. 27-36.

Research output: Contribution to journalArticle

Luria, Gianvittorio ; Duran, Dunja ; Visani, Elisa ; Sommariva, Sara ; Rotondi, Fabio ; Rossi Sebastiano, Davide ; Panzica, Ferruccio ; Piana, Michele ; Sorrentino, Alberto. / Bayesian multi-dipole modelling in the frequency domain. In: Journal of Neuroscience Methods. 2019 ; Vol. 312. pp. 27-36.
@article{da2691d0f06a445196d4f080ec679fb9,
title = "Bayesian multi-dipole modelling in the frequency domain",
abstract = "Background: Magneto- and Electro-encephalography record the electromagnetic field generated by neural currents with high temporal frequency and good spatial resolution, and are therefore well suited for source localization in the time and in the frequency domain. In particular, localization of the generators of neural oscillations is very important in the study of cognitive processes in the healthy and in the pathological brain. New method: We introduce the use of a Bayesian multi-dipole localization method in the frequency domain. Given the Fourier Transform of the data at one or multiple frequencies and/or trials, the algorithm approximates numerically the posterior distribution with Monte Carlo techniques. Results: We use synthetic data to show that the proposed method behaves well under a wide range of experimental conditions, including low signal-to-noise ratios and correlated sources. We use dipole clusters to mimic the effect of extended sources. In addition, we test the algorithm on real MEG data to confirm its feasibility. Comparison with existing method(s): Throughout the whole study, DICS (Dynamic Imaging of Coherent Sources) is used systematically as a benchmark. The two methods provide similar general pictures; the posterior distributions of the Bayesian approach contain much richer information at the price of a higher computational cost. Conclusions: The Bayesian method described in this paper represents a reliable approach for localization of multiple dipoles in the frequency domain.",
keywords = "Bayesian methods, EEG/MEG, Oscillatory brain activity, Sequential Monte Carlo, Source modelling",
author = "Gianvittorio Luria and Dunja Duran and Elisa Visani and Sara Sommariva and Fabio Rotondi and {Rossi Sebastiano}, Davide and Ferruccio Panzica and Michele Piana and Alberto Sorrentino",
year = "2019",
month = "1",
day = "15",
doi = "10.1016/j.jneumeth.2018.11.007",
language = "English",
volume = "312",
pages = "27--36",
journal = "Journal of Neuroscience Methods",
issn = "0165-0270",
publisher = "Elsevier",

}

TY - JOUR

T1 - Bayesian multi-dipole modelling in the frequency domain

AU - Luria, Gianvittorio

AU - Duran, Dunja

AU - Visani, Elisa

AU - Sommariva, Sara

AU - Rotondi, Fabio

AU - Rossi Sebastiano, Davide

AU - Panzica, Ferruccio

AU - Piana, Michele

AU - Sorrentino, Alberto

PY - 2019/1/15

Y1 - 2019/1/15

N2 - Background: Magneto- and Electro-encephalography record the electromagnetic field generated by neural currents with high temporal frequency and good spatial resolution, and are therefore well suited for source localization in the time and in the frequency domain. In particular, localization of the generators of neural oscillations is very important in the study of cognitive processes in the healthy and in the pathological brain. New method: We introduce the use of a Bayesian multi-dipole localization method in the frequency domain. Given the Fourier Transform of the data at one or multiple frequencies and/or trials, the algorithm approximates numerically the posterior distribution with Monte Carlo techniques. Results: We use synthetic data to show that the proposed method behaves well under a wide range of experimental conditions, including low signal-to-noise ratios and correlated sources. We use dipole clusters to mimic the effect of extended sources. In addition, we test the algorithm on real MEG data to confirm its feasibility. Comparison with existing method(s): Throughout the whole study, DICS (Dynamic Imaging of Coherent Sources) is used systematically as a benchmark. The two methods provide similar general pictures; the posterior distributions of the Bayesian approach contain much richer information at the price of a higher computational cost. Conclusions: The Bayesian method described in this paper represents a reliable approach for localization of multiple dipoles in the frequency domain.

AB - Background: Magneto- and Electro-encephalography record the electromagnetic field generated by neural currents with high temporal frequency and good spatial resolution, and are therefore well suited for source localization in the time and in the frequency domain. In particular, localization of the generators of neural oscillations is very important in the study of cognitive processes in the healthy and in the pathological brain. New method: We introduce the use of a Bayesian multi-dipole localization method in the frequency domain. Given the Fourier Transform of the data at one or multiple frequencies and/or trials, the algorithm approximates numerically the posterior distribution with Monte Carlo techniques. Results: We use synthetic data to show that the proposed method behaves well under a wide range of experimental conditions, including low signal-to-noise ratios and correlated sources. We use dipole clusters to mimic the effect of extended sources. In addition, we test the algorithm on real MEG data to confirm its feasibility. Comparison with existing method(s): Throughout the whole study, DICS (Dynamic Imaging of Coherent Sources) is used systematically as a benchmark. The two methods provide similar general pictures; the posterior distributions of the Bayesian approach contain much richer information at the price of a higher computational cost. Conclusions: The Bayesian method described in this paper represents a reliable approach for localization of multiple dipoles in the frequency domain.

KW - Bayesian methods

KW - EEG/MEG

KW - Oscillatory brain activity

KW - Sequential Monte Carlo

KW - Source modelling

UR - http://www.scopus.com/inward/record.url?scp=85057014395&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85057014395&partnerID=8YFLogxK

U2 - 10.1016/j.jneumeth.2018.11.007

DO - 10.1016/j.jneumeth.2018.11.007

M3 - Article

C2 - 30452978

AN - SCOPUS:85057014395

VL - 312

SP - 27

EP - 36

JO - Journal of Neuroscience Methods

JF - Journal of Neuroscience Methods

SN - 0165-0270

ER -