Characterization of Event Related Potentials Using Information Theoretic Distance Measures

Abstract

Analysis of event-related potentials (ERPs) using signal processing tools has become extremely widespread in recent years. Nonstationary signal processing tools such as wavelets and time-frequency distributions have proven to be especially effective in characterizing the transient phenomena encountered in event-related potentials. In this paper, we focus on the analysis of event-related potentials collected during a psychological experiment where two groups of subjects, spider phobics and snake phobics, are shown the same set of stimulus: A blank stimulus, a neutral stimulus and a spider stimulus. We introduce a new approach, based on time-frequency distributions, for analyzing the ERPs. The difference in brain activity before and after a stimulus is presented is quantified using distance measures as adapted to the time-frequency plane. Three different distance measures, including a new information theoretic distance measure, are applied on the time-frequency plane to discriminate between the responses of the two groups of subjects. The results illustrate the effectiveness of using distance measures combined with time-frequency distributions in differentiating between the two classes of subjects and the different regions of the brain.

Regional Analysis of Spontaneous MEG Rhythms in Patients with Alzheimer’s Disease Using Spectral Entropies

“…The PSD n represents again the probability density function, analogously to previous EEG studies. 2,24 Hence, the Re´nyi entropy can be considered as an alternative way to estimate the irregularity of timefrequency distributions. 3 This entropic form is parameterized by an entropic index q 2 <, in order that the Re´nyi entropy can be reduced to the BoltzmannGibbs entropy in the limit q fi 1.…”

“…Previous studies have successfully applied several quantifiers based on extensive (e.g., Shannon and Re´nyi entropies) and non-extensive (e.g., Tsallis entropy) information measures to analyze EEG signals. 2,5,9,24,33,40,41 While extensive quantifiers, such as Shannon and Re´nyi entropies, have proved successful in describing systems with short-range interactions, non-extensive measures, such as Tsallis entropy, are able to describe a system where the effective interactions are of long-range. 5,40 In addition, MEG recordings, like EEG signals, are non-stationary and their characteristics may change over time.…”

“…Previous efforts have exploited the analogy between power spectral densities and probability densities to apply several entropic forms to time-frequency distributions. 2,3 The present study is a new approach to explore the ability of several measures derived from information theory to characterize MEG rhythms in AD. This work has been developed on the basis of the results obtained in a pilot study where we tested several entropies using the averaged relative power over all sensors as a probability distribution.…”

Information Theoretic Measures for Quantifying the Integration of Neural Activity

“…The entropy of the resulting TFD surface is computed using Rényi entropy of order 3. The choice of the order of the Rényi entropy has been previously discussed in detail in [13]. The entropy values are normalized by log 2 L 2 = log 2 2500.…”

“…We use recent results in information-theoretic analysis of time-frequency distributions [12], [13] to define entropy and divergence functions based on this joint density function. Second, we extend the current work by considering the interactions between different electrodes as well as the complexity of the individual signals.…”