Limits...
Maximum-likelihood estimation of channel-dependent trial-to-trial variability of auditory evoked brain responses in MEG.

Sielużycki C, Kordowski P - Biomed Eng Online (2014)

Bottom Line: Following the work of de Munck et al., our approach is based on the maximum likelihood estimation and involves an approximation of the spatio-temporal covariance of the contaminating background noise by means of the Kronecker product of its spatial and temporal covariance matrices.We also present an illustrative example of the application of this methodology to real MEG data taken from an auditory experimental paradigm, where we found hemispheric lateralization of the habituation effect to multiple stimulus presentation.Hence, it may be a useful tool in paradigms that assume lateralization effects, like, e.g., those involving language processing.

View Article: PubMed Central - HTML - PubMed

Affiliation: Special Lab Non-invasive Brain Imaging, Leibniz Institute for Neurobiology, Brenneckestr, 6, 39118 Magdeburg, Germany. cezary.sieluzycki@icm-institute.org.

ABSTRACT

Background: We propose a mathematical model for multichannel assessment of the trial-to-trial variability of auditory evoked brain responses in magnetoencephalography (MEG).

Methods: Following the work of de Munck et al., our approach is based on the maximum likelihood estimation and involves an approximation of the spatio-temporal covariance of the contaminating background noise by means of the Kronecker product of its spatial and temporal covariance matrices. Extending the work of de Munck et al., where the trial-to-trial variability of the responses was considered identical to all channels, we evaluate it for each individual channel.

Results: Simulations with two equivalent current dipoles (ECDs) with different trial-to-trial variability, one seeded in each of the auditory cortices, were used to study the applicability of the proposed methodology on the sensor level and revealed spatial selectivity of the trial-to-trial estimates. In addition, we simulated a scenario with neighboring ECDs, to show limitations of the method. We also present an illustrative example of the application of this methodology to real MEG data taken from an auditory experimental paradigm, where we found hemispheric lateralization of the habituation effect to multiple stimulus presentation.

Conclusions: The proposed algorithm is capable of reconstructing lateralization effects of the trial-to-trial variability of evoked responses, i.e. when an ECD of only one hemisphere habituates, whereas the activity of the other hemisphere is not subject to habituation. Hence, it may be a useful tool in paradigms that assume lateralization effects, like, e.g., those involving language processing.

Show MeSH

Related in: MedlinePlus

Simulated ECD locations. Like in figure 1 but with the additional (third) ECD in the right auditory cortex.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4060856&req=5

Figure 6: Simulated ECD locations. Like in figure 1 but with the additional (third) ECD in the right auditory cortex.

Mentions: In order to test the performance of the method in situations in which the leadfields of the ECDs are far from mutual orthogonality, we simulated an additional scenario in which a third ECD was placed in the right hemisphere; see figures 6 and 7. The time course of this third ECD was simulated as s3(j) = sin(3j/J + π/15) normalized with respect to the peak value. Due to two ECDs in the right hemisphere in this scenario, the overall ECD signal strength in that hemisphere was larger, compared to that in the left hemisphere. Moreover, due to the aforementioned scaling of the signal and the noise resulting in an overall SNR of 0.1, the SNR for the left-hemisphere sensors was now weaker than in the previous simulations. To examine the influence of the trial-to-trial characteristics of the third source on the overall habituation pattern for the right hemisphere, we assumed that this additional source did not habituate, i.e. that its h3(k) = 1 for all k (h1 and h2 were like in the first simulation; see figure 2). Since, as aforementioned, ECDs whose leadfields are not distinct enough, i.e. not orthogonal or almost orthogonal, can not be efficiently separated, the estimated overall habituation pattern for the right hemisphere is weakened in this scenario; see figure 8. This is due to the fact that the third source did not habituate. The observed pattern is a mixture of the trial-to-trial characteristics of the two sources in the right hemisphere; the mean of the significantly non-zero slopes of the 1st-degree polynomials fitted to the estimates of ψ for sensors above the right-hemisphere was equal to -0.0036 (with σ=0.0006). This will of course be the case for any real-data scenario with multiple sources located close to each other. However, since the exact number of ECDs is not known in practice, we do not consider such an “integrative”, and hence impaired, resolution of the method a serious flaw. The fact that we do not know, and hence do not want to assume explicitly, the exact number of ECDs for real data is essentially the reason for estimating the trial-to-trial variability on the sensor rather than source level.


Maximum-likelihood estimation of channel-dependent trial-to-trial variability of auditory evoked brain responses in MEG.

Sielużycki C, Kordowski P - Biomed Eng Online (2014)

Simulated ECD locations. Like in figure 1 but with the additional (third) ECD in the right auditory cortex.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4060856&req=5

Figure 6: Simulated ECD locations. Like in figure 1 but with the additional (third) ECD in the right auditory cortex.
Mentions: In order to test the performance of the method in situations in which the leadfields of the ECDs are far from mutual orthogonality, we simulated an additional scenario in which a third ECD was placed in the right hemisphere; see figures 6 and 7. The time course of this third ECD was simulated as s3(j) = sin(3j/J + π/15) normalized with respect to the peak value. Due to two ECDs in the right hemisphere in this scenario, the overall ECD signal strength in that hemisphere was larger, compared to that in the left hemisphere. Moreover, due to the aforementioned scaling of the signal and the noise resulting in an overall SNR of 0.1, the SNR for the left-hemisphere sensors was now weaker than in the previous simulations. To examine the influence of the trial-to-trial characteristics of the third source on the overall habituation pattern for the right hemisphere, we assumed that this additional source did not habituate, i.e. that its h3(k) = 1 for all k (h1 and h2 were like in the first simulation; see figure 2). Since, as aforementioned, ECDs whose leadfields are not distinct enough, i.e. not orthogonal or almost orthogonal, can not be efficiently separated, the estimated overall habituation pattern for the right hemisphere is weakened in this scenario; see figure 8. This is due to the fact that the third source did not habituate. The observed pattern is a mixture of the trial-to-trial characteristics of the two sources in the right hemisphere; the mean of the significantly non-zero slopes of the 1st-degree polynomials fitted to the estimates of ψ for sensors above the right-hemisphere was equal to -0.0036 (with σ=0.0006). This will of course be the case for any real-data scenario with multiple sources located close to each other. However, since the exact number of ECDs is not known in practice, we do not consider such an “integrative”, and hence impaired, resolution of the method a serious flaw. The fact that we do not know, and hence do not want to assume explicitly, the exact number of ECDs for real data is essentially the reason for estimating the trial-to-trial variability on the sensor rather than source level.

Bottom Line: Following the work of de Munck et al., our approach is based on the maximum likelihood estimation and involves an approximation of the spatio-temporal covariance of the contaminating background noise by means of the Kronecker product of its spatial and temporal covariance matrices.We also present an illustrative example of the application of this methodology to real MEG data taken from an auditory experimental paradigm, where we found hemispheric lateralization of the habituation effect to multiple stimulus presentation.Hence, it may be a useful tool in paradigms that assume lateralization effects, like, e.g., those involving language processing.

View Article: PubMed Central - HTML - PubMed

Affiliation: Special Lab Non-invasive Brain Imaging, Leibniz Institute for Neurobiology, Brenneckestr, 6, 39118 Magdeburg, Germany. cezary.sieluzycki@icm-institute.org.

ABSTRACT

Background: We propose a mathematical model for multichannel assessment of the trial-to-trial variability of auditory evoked brain responses in magnetoencephalography (MEG).

Methods: Following the work of de Munck et al., our approach is based on the maximum likelihood estimation and involves an approximation of the spatio-temporal covariance of the contaminating background noise by means of the Kronecker product of its spatial and temporal covariance matrices. Extending the work of de Munck et al., where the trial-to-trial variability of the responses was considered identical to all channels, we evaluate it for each individual channel.

Results: Simulations with two equivalent current dipoles (ECDs) with different trial-to-trial variability, one seeded in each of the auditory cortices, were used to study the applicability of the proposed methodology on the sensor level and revealed spatial selectivity of the trial-to-trial estimates. In addition, we simulated a scenario with neighboring ECDs, to show limitations of the method. We also present an illustrative example of the application of this methodology to real MEG data taken from an auditory experimental paradigm, where we found hemispheric lateralization of the habituation effect to multiple stimulus presentation.

Conclusions: The proposed algorithm is capable of reconstructing lateralization effects of the trial-to-trial variability of evoked responses, i.e. when an ECD of only one hemisphere habituates, whereas the activity of the other hemisphere is not subject to habituation. Hence, it may be a useful tool in paradigms that assume lateralization effects, like, e.g., those involving language processing.

Show MeSH
Related in: MedlinePlus