Inferring Brain Signals Synchronicity from a Sample of EEG Readings - PubMed (original) (raw)

Inferring Brain Signals Synchronicity from a Sample of EEG Readings

Qian Li et al. J Am Stat Assoc. 2019.

Abstract

Inferring patterns of synchronous brain activity from a heterogeneous sample of electroencephalograms (EEG) is scientifically and methodologically challenging. While it is intuitively and statistically appealing to rely on readings from more than one individual in order to highlight recurrent patterns of brain activation, pooling information across subjects presents non-trivial methodological problems. We discuss some of the scientific issues associated with the understanding of synchronized neuronal activity and propose a methodological framework for statistical inference from a sample of EEG readings. Our work builds on classical contributions in time-series, clustering and functional data analysis, in an effort to reframe a challenging inferential problem in the context of familiar analytical techniques. Some attention is paid to computational issues, with a proposal based on the combination of machine learning and Bayesian techniques.

Keywords: Consensus Clustering; EEG; Hierarchical Mixture Models; Spectral Clustering.

PubMed Disclaimer

Figures

Figure 1

Figure 1. Simulated spectral configurations

(a) main-state spectral densities. (b) off-state spectral densities. (c) Segment-by-segment normalized power spectral densities for a piecewise stationary process simulated from cluster 4.

Figure 2

Figure 2. Simulation results

(a) Path-length for the search in Algorithm 1 for varying smoothing configurations in γ. (b) Estimated adherence parameters α̂ ’s and 95% credible intervals against the data generating truth. (c) Clustering accuracy against generating _α_’s at the subject-and population-level. (d) Average difference in clustering variance against true _α_’s.

Figure 3

Figure 3. Synchronicity and spectral features

For each cohort, cluster configurations are depicted for two illustrative subjects. For each electrode, the estimated spectral density (normalized) is color coded by cluster membership. All plots refer to the epoch that is most coherent with subject-level clustering.

Figure 4

Figure 4. Group contrasts, ASD (1) vs TD (2)

(1.a) TD-cohort posterior least square synchronicity. (1.b) TD-cohort normalized posterior entropy. (1.c) TD-cohort subject- and population-level cluster assignments. (2.a) ASD-cohort posterior least square synchronicity. (2.b) ASD-cohort normalized posterior entropy. (2.c) ASD-cohort subject- and population-level cluster assignments. In the (c) panels, we report consensus labels as the last row. Subject-level labels are reported in each row, together with posterior median estimates of cluster adherence.

References

    1. Bastos AM, Schoffelen J. A Tutorial Review of functional connectivity analysis methods and their interpretation pitfalls. Frontiers in System Neuroscience. 2016;9:175. -PMC -PubMed
    1. Bigdely-Shamlo N, Mullen T, Kothe C, Su K-M, Robbins KA. The PREP pipeline: standardized preprocessing for large-scale EEG analysis. Frontiers in Neuroinformatics. 2015;9:16. URL: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4471356/ -PMC -PubMed
    1. Broyd S, Demanuele C, Debener S, Helps K, James C, Sonuga-Barke EJ. Default-mode brain dysfunction in mental disorders: a systematic review. Nature Biobehavioral Review. 2008;33(3):279–296. -PubMed
    1. Burnette CP, Henderson HA, Inge AP, Zahka NE, Schwartz CB, Mundy PC. Anterior EEG asymmetry and the modifier model of autism. Journal of autism and developmental disorders. 2011;41(8):1113–1124. -PMC -PubMed
    1. Celeux G, Hurn M, Robert CP. Computational and inferential difficulties with mixture posterior distributions. Journal of the American Statistical Association. 2000;95(451):957–970.

LinkOut - more resources