Biomag 96 pp 278-281 | Cite as

Development, Implementation and Applications of Single Epoch Analysis of MEG Signals

  • L. C. Liu
  • A. A. Ioannides
  • H. Heer
  • M. Wagener
  • F. Delonge
  • H. Halling
  • H. W. Müller-Gärtner
Conference paper

Abstract

In both MEG and EEG studies, single epoch analysis has become increasingly populär [1, 2]. In EEG the relationship between generators and signal is blurred by the high resistivity and inhomogeneity of the skull, making single trial analysis problematic because of low signal-to-noise ratio. For MEG, the signal-to-noise ratio in single trials is high, in the sense that, contributions from small cortical regions are easily detectable, although not easily distinguished from other contributions. In addition, single trial analysis of MEG signals incurs a high computational bürden and produces huge amount of output. In our earlier work [3, 4], we have addressed these problems by introducing a vector signal transformation called V3, which is the MEG analogue of the well-known Laplacian method of EEG. We have tested the accuracy of the V3 based estimates with Computer generated data and compared the results with the ones obtained from Magnetic Field Tomography (MFT) analysis of the same data. These tests have shown that the V3 based estimates are remarkably accurate when superficial generators are involved [4]. Estimators constructed from correlational measures based on the V3 and integrated over a period of time and/or a region, can be used to scan quickly and efficiently through all single trials. The V3 analysis has already been applied to the identification and quantification of auditory cortex activation and the investigation of gamma-band activity in single trials [4, 5]. Further correlation analysis of spatio-temporal V3 templates, has enabled us to describe qualitatively and quantitatively how patterns observed in the average signal are represented in single trials [6]: we showed conclusively that the average is a superposition of histories each reflected in different subsets of the epochs. In this paper, we will show another related application of the V3: it can be used as the basis for easy and fast inspection of data, and for identification of correlations in activity, in space and time, within the same hemisphere or across the hemispheres.

Keywords

Milton Alho 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    A.R. Haig, E. Gordon, G. Rogers, and J. Anderson. ‘Classification of Single-trial ERP Sub-types: Application of Globally Optimal Vector Quantization using Simulated Annealing’. Electroenceph. Clin. Neurophysiol., 94: 288–297, 1995.CrossRefGoogle Scholar
  2. [2]
    B. Lütkenhöner, C. Pantev, A. Grunwald, and E. Menninghaus. ‘Source-space Projection of Single Trials of the Auditory Evoked Field’. In C. Baumgartner, L. Deeke, G. Stroink, and S.J. Williamson, editors, ’Biomagnetism: Fundamental Research and Clinical Applications’, pages 347–351. Elsevier Science, Amsterdam, 1995.Google Scholar
  3. A.A. Ioannides, R. Hasson, and G.J. Miseldine. ‘Model-dependent Noise Elimination and Distributed Source Solutions for the Biomagnetic Inverse Problem’. In A.F. Gmitro and et al., editors, ’Digital Image Synthesis and Inverse Optics\ pages 471–481. Proc. SPIE 1351, 1990.Google Scholar
  4. [4]
    L.C. Liu. ‘Single Epoch Analysis and Bi-hemispheric Study of Magnetoencephalographic (MEG) Signals using Vector Signal Transformation V3 and Magnetic Field Tomography (MFT)’. PhD thesis, Department of Physics, The Open University, U.K., February 1995.Google Scholar
  5. [5]
    L.C. Liu and A.A. Ioannides. ‘Single Epoch Analysis of MEG Signals’. In C. Baumgartner, L. Deeke, G. Stroink, and S.J. Williamson, editors, ’Biomagnetism: Fundamental Research and Clinical Applications pages 439–444. Elsevier Science, Amsterdam, 1995.Google Scholar
  6. [6]
    L.C. Liu and A.A. Ioannides. ‘A Correlation Study of Averaged and Single Trial MEG Signals: The Average Describes Multiple Histories Each in a Different Set of Single Trials’. Brain Topography, 1996. Accepted for publication.Google Scholar
  7. [7]
    M. Huotilainen, R.J. Ilmoniemi, H. Tiitinen, J. Lavikainen, K. Alho, M. Kajola and R. Näatänen. ‘The Projection Method in Removing Eye-blink Artefacts from Multichannel MEG Measurements1. In C. Baumgartner, L. Deeke, G. Stroink, and S.J. Williamson, editors, ’Biomagnetism: Fundamental Research and Clinical Applicationspages 363–367. Elsevier Science, Amsterdam, 1995.Google Scholar

Copyright information

© Springer Science+Business Media New York 2000

Authors and Affiliations

  • L. C. Liu
    • 1
  • A. A. Ioannides
    • 1
    • 3
  • H. Heer
    • 2
  • M. Wagener
    • 2
  • F. Delonge
    • 2
  • H. Halling
    • 2
  • H. W. Müller-Gärtner
    • 1
  1. 1.Institute of MedicineJülichGermany
  2. 2.ElectronicsResearch Centre JülichJülichGermany
  3. 3.Department of PhysicsThe Open UniversityMilton KeynesUK

Personalised recommendations