Conclusion

Our objective to explore the clinical usefulness of EEG data as a source of biomarkers was accomplished by through the methods of network and time-series analysis. Clear distinctions were observed between the control and test subject sets under both methods of analysis. While the control and test groups shared similarities in the spatial network interactions, network statistics were calculated to capture key differences in the patterns of connectivity between the two groups. The observations made using the three methods of time-series analysis, threshold, autocorrelation, and wavelet, confirmed that there exist key differences in the time-evolution characteristics of the control and test groups. Despite the time and resource constraints that limited the goals in this project, further research can be conducted to model more complex or nonlinear interactions in the brain. Given that the time-series portion of our analysis was purely exploratory, preliminary observations should be further grounded by the neurophysiological context of the findings, which was beyond the scope of this project. Furthermore, in future work, using alternate methods of data processing can address the effects of volume conduction, field spread, and artifacts in the data. Amidst the challenges we faced as a result of our limited background, our primary objectives were met and the results gathered confirm the potential of EEG data analysis to be utilized as a non-invasive diagnostic tool.