Uncertainty in Functional Network Representations of Brain Activity of Alcoholic Patients

Abstract

In spite of the large attention received by brain activity analyses through functional networks, the effects of uncertainty on such representations have mostly been neglected. We here elaborate the hypothesis that such uncertainty is not just a nuisance, but that on the contrary is condition-dependent.We test this hypothesis by analysing a large set of EEG brain recordings corresponding to control subjects and patients suffering from alcoholism, through the reconstruction of the corresponding Maximum Spanning Trees (MSTs), the assessment of their topological differences, and the comparison of two frequenstist and Bayesian reconstruction approaches. A machine learning model

Algorithmic Approaches for Assessing Irreversibility in Time Series: Review and Comparison

“…If the analysis of synthetic data depicted a complex scenario, things only get more interesting and complex when moving to real-world data. As a prototypical example, we here present an irreversibility analysis of human electro-encephalography (EEG) data representing the brain activity of control subjects and patients suffering from alcoholism [122][123][124] and freely available at https://archive.ics.uci.edu/ml/datasets/EEG+Database, accessed on 4 November 2021. Each recording corresponds to the execution of a standard object recognition task [125], and includes 64 time series (i.e., one for each of the 64 electrodes or channels) of 256 elements (i.e., one second of brain activity recorded at 256 Hz).…”

“…Each recording corresponds to the execution of a standard object recognition task [125], and includes 64 time series (i.e., one for each of the 64 electrodes or channels) of 256 elements (i.e., one second of brain activity recorded at 256 Hz). Note that the assessment of the irreversibility on this data set is a challenging task, as, firstly, time series are intrinsically noisy due to the technical limitations of EEG recordings; and secondly, they are short, following brain non-stationarity [124]. On the other hand, a large number of trials are available in both groups (respectively 2010 and 3467 for control subjects and patients), thus enabling robust statistical results.…”

Assessing time series irreversibility through micro-scale trends

“…As a first real-world test, we here apply the proposed irreversibility metric to a data set of electroencephalographic (EEG) recordings, comprising both control subjects and patients suffering from alcoholism [38][39][40] and available at https://archive.ics.uci.edu/ml/ datasets/EEG+Database. Each recording corresponds to the execution of a standard object recognition task 41 , and includes 64 time series (i.e.…”