Estimating Directed Phase-Amplitude Interactions from EEG Data through Kernel-Based Phase Transfer Entropy

Iván De La Pava Panche; Viviana Gómez-Orozco; Andrés Álvarez-Meza; David Cárdenas-Peña; Álvaro Orozco-Gutiérrez

doi:10.3390/app11219803

Estimating Directed Phase-Amplitude Interactions from EEG Data through Kernel-Based Phase Transfer Entropy

Applied Sciences ◽

10.3390/app11219803 ◽

2021 ◽

Vol 11 (21) ◽

pp. 9803

Author(s):

Iván De La Pava Panche ◽

Viviana Gómez-Orozco ◽

Andrés Álvarez-Meza ◽

David Cárdenas-Peña ◽

Álvaro Orozco-Gutiérrez

Keyword(s):

Working Memory ◽

Time Series ◽

Cognitive Task ◽

Memory Task ◽

Synthetic Data ◽

Transfer Entropy ◽

Alternative Methods ◽

Interaction Detection ◽

Eeg Data ◽

Phase Amplitude

Cross-frequency interactions, a form of oscillatory neural activity, are thought to play an essential role in the integration of distributed information in the brain. Indeed, phase-amplitude interactions are believed to allow for the transfer of information from large-scale brain networks, oscillating at low frequencies, to local, rapidly oscillating neural assemblies. A promising approach to estimating such interactions is the use of transfer entropy (TE), a non-linear, information-theory-based effective connectivity measure. The conventional method involves feeding instantaneous phase and amplitude time series, extracted at the target frequencies, to a TE estimator. In this work, we propose that the problem of directed phase-amplitude interaction detection is recast as a phase TE estimation problem, under the hypothesis that estimating TE from data of the same nature, i.e., two phase time series, will improve the robustness to the common confounding factors that affect connectivity measures, such as the presence of high noise levels. We implement our proposal using a kernel-based TE estimator, defined in terms of Renyi’s α entropy, which has successfully been used to compute single-trial phase TE. We tested our approach on the synthetic data generated through a simulation model capable of producing a time series with directed phase-amplitude interactions at two given frequencies, and on EEG data from a cognitive task designed to activate working memory, a memory system whose underpinning mechanisms are thought to include phase–amplitude couplings. Our proposal detected statistically significant interactions between the simulated signals at the desired frequencies for the synthetic data, identifying the correct direction of the interaction. It also displayed higher robustness to noise than the alternative methods. The results attained for the working memory data showed that the proposed approach codes connectivity patterns based on directed phase–amplitude interactions, that allow for the different cognitive load levels of the working memory task to be differentiated.

Download Full-text

Negative Effects of Embodiment in a Visuo-Spatial Working Memory Task in Children, Young Adults, and Older Adults

Frontiers in Psychology ◽

10.3389/fpsyg.2021.688174 ◽

2021 ◽

Vol 12 ◽

Author(s):

Gianluca Amico ◽

Sabine Schaefer

Keyword(s):

Older Adults ◽

Working Memory ◽

Young Adults ◽

Spatial Information ◽

Spatial Working Memory ◽

Cognitive Task ◽

Memory Performance ◽

Age Groups ◽

Memory Task ◽

Negative Effects

Studies examining the effect of embodied cognition have shown that linking one’s body movements to a cognitive task can enhance performance. The current study investigated whether concurrent walking while encoding or recalling spatial information improves working memory performance, and whether 10-year-old children, young adults, or older adults (Mage = 72 years) are affected differently by embodiment. The goal of the Spatial Memory Task was to encode and recall sequences of increasing length by reproducing positions of target fields in the correct order. The nine targets were positioned in a random configuration on a large square carpet (2.5 m × 2.5 m). During encoding and recall, participants either did not move, or they walked into the target fields. In a within-subjects design, all possible combinations of encoding and recall conditions were tested in counterbalanced order. Contrary to our predictions, moving particularly impaired encoding, but also recall. These negative effects were present in all age groups, but older adults’ memory was hampered even more strongly by walking during encoding and recall. Our results indicate that embodiment may not help people to memorize spatial information, but can create a dual-task situation instead.

Download Full-text

The Role of Physical Fitness in Cognitive-Related Biomarkers in Persons at Genetic Risk of Familial Alzheimer’s Disease

Journal of Clinical Medicine ◽

10.3390/jcm8101639 ◽

2019 ◽

Vol 8 (10) ◽

pp. 1639 ◽

Cited By ~ 3

Author(s):

Tsai ◽

Sun ◽

Kuo ◽

Pai

Keyword(s):

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Working Memory ◽

Physical Fitness ◽

Cognitive Task ◽

Apoe Genotype ◽

Memory Task ◽

Molecular Biomarkers ◽

Interactive Effects ◽

Visuospatial Working Memory

Introduction: Nondemented people with a family history of Alzheimer’s disease (ADFH) and the ApoE-4 allele have been demonstrated to show a trend for a higher probability of cognitive decline and aberrant levels of cognitive-related biomarkers. However, the potential interactive effects on physical fitness have not been investigated. Purpose: The primary purpose of this study was to determine whether ADFH individuals with the ApoE-4 genotype show deviant brain event-related neural oscillatory performance and cognitively-related molecular indices. A secondary purpose was to examine the interactive effects on physical fitness. Methods: Blood samples were provided from 110 individuals with ADFH to assess molecular biomarkers and the ApoE genotype for the purpose of dividing them into an ApoE-4 group (n = 16) and a non-ApoE-4 group (n = 16) in order for them to complete a visuospatial working memory task while simultaneously recording electroencephalographic signals. They also performed a senior functional physical fitness (SFPF) test. Results: While performing the cognitive task, the ApoE-4 relative to non-ApoE-4 group showed worse accuracy rates (ARs) and brain neural oscillatory performance. There were no significant between-group differences with regard to any molecular biomarkers (e.g., IL-1β, IL-6, IL-8, BDNF, Aβ1-40, Aβ1-42). VO2max was significantly correlated with the neuropsychological performance (i.e., ARs and RTs) in the 2-item and 4-item conditions in the ApoE-4 group and across the two groups. However, the electroencephalogram (EEG) oscillations during visuospatial working memory processing in the two conditions were not correlated with any SFPF scores or cardiorespiratory tests in the two groups. Conclusions: ADFH individuals with the ApoE-4 genotype only showed deviant neuropsychological (e.g., ARs) and neural oscillatory performance when performing the cognitive task with a higher visuospatial working memory load. Cardiorespiratory fitness potentially played an important role in neuropsychological impairment in this group.

Download Full-text

Information-theoretic measures for nonlinear causality detection: application to social media sentiment and cryptocurrency prices

Royal Society Open Science ◽

10.1098/rsos.200863 ◽

2020 ◽

Vol 7 (9) ◽

pp. 200863

Author(s):

Z. Keskin ◽

T. Aste

Keyword(s):

Time Series ◽

Information Transfer ◽

Synthetic Data ◽

Transfer Entropy ◽

Significant Information ◽

Information Theoretic ◽

Order Of Magnitude ◽

Information Theoretic Measures ◽

Media Sentiment ◽

Non Parametric

Information transfer between time series is calculated using the asymmetric information-theoretic measure known as transfer entropy. Geweke’s autoregressive formulation of Granger causality is used to compute linear transfer entropy, and Schreiber’s general, non-parametric, information-theoretic formulation is used to quantify nonlinear transfer entropy. We first validate these measures against synthetic data. Then we apply these measures to detect statistical causality between social sentiment changes and cryptocurrency returns. We validate results by performing permutation tests by shuffling the time series, and calculate the Z -score. We also investigate different approaches for partitioning in non-parametric density estimation which can improve the significance. Using these techniques on sentiment and price data over a 48-month period to August 2018, for four major cryptocurrencies, namely bitcoin (BTC), ripple (XRP), litecoin (LTC) and ethereum (ETH), we detect significant information transfer, on hourly timescales, with greater net information transfer from sentiment to price for XRP and LTC, and instead from price to sentiment for BTC and ETH. We report the scale of nonlinear statistical causality to be an order of magnitude larger than the linear case.

Download Full-text

Kernel-Based Phase Transfer Entropy with Enhanced Feature Relevance Analysis for Brain Computer Interfaces

Applied Sciences ◽

10.3390/app11156689 ◽

2021 ◽

Vol 11 (15) ◽

pp. 6689

Author(s):

Iván De La Pava Panche ◽

Andrés Álvarez-Meza ◽

Paula Marcela Herrera Gómez ◽

David Cárdenas-Peña ◽

Jorge Iván Ríos Patiño ◽

...

Keyword(s):

Working Memory ◽

Time Series ◽

Motor Imagery ◽

Phase Transfer ◽

Effective Connectivity ◽

Transfer Entropy ◽

Neural Oscillations ◽

Computer Interfaces ◽

Relevance Analysis ◽

The Brain

Neural oscillations are present in the brain at different spatial and temporal scales, and they are linked to several cognitive functions. Furthermore, the information carried by their phases is fundamental for the coordination of anatomically distributed processing in the brain. The concept of phase transfer entropy refers to an information theory-based measure of directed connectivity among neural oscillations that allows studying such distributed processes. Phase TE is commonly obtained from probability estimations carried out over data from multiple trials, which bars its use as a characterization strategy in brain–computer interfaces. In this work, we propose a novel methodology to estimate TE between single pairs of instantaneous phase time series. Our approach combines a kernel-based TE estimator defined in terms of Renyi’s α entropy, which sidesteps the need for probability distribution computation with phase time series obtained by complex filtering the neural signals. Besides, a kernel-alignment-based relevance analysis is added to highlight relevant features from effective connectivity-based representation supporting further classification stages in EEG-based brain–computer interface systems. Our proposal is tested on simulated coupled data and two publicly available databases containing EEG signals recorded under motor imagery and visual working memory paradigms. Attained results demonstrate how the introduced effective connectivity succeeds in detecting the interactions present in the data for the former, with statistically significant results around the frequencies of interest. It also reflects differences in coupling strength, is robust to realistic noise and signal mixing levels, and captures bidirectional interactions of localized frequency content. Obtained results for the motor imagery and working memory databases show that our approach, combined with the relevance analysis strategy, codes discriminant spatial and frequency-dependent patterns for the different conditions in each experimental paradigm, with classification performances that do well in comparison with those of alternative methods of similar nature.

Download Full-text

Heart rate variability change during a stressful cognitive task in individuals with anxiety and control participants

BMC Psychology ◽

10.1186/s40359-021-00551-4 ◽

2021 ◽

Vol 9 (1) ◽

Author(s):

Judith Held ◽

Andreea Vîslă ◽

Christine Wolfer ◽

Nadine Messerli-Bürgy ◽

Christoph Flückiger

Keyword(s):

Heart Rate ◽

Working Memory ◽

Heart Rate Variability ◽

Cognitive Task ◽

Memory Task ◽

Patient Characteristics ◽

Recovery Phase ◽

Control Group ◽

Anxious Group ◽

And Control

Abstract Background Individuals suffering from an anxiety disorder are characterized by chronically low heart rate variability (HRV) compared to healthy individuals during resting state conditions. However, when examining HRV and HR in response to a stressor, mixed results have been obtained when comparing anxious and non-anxious groups. Methods The primary aim of the present study was to investigate HRV and HR responding in 26 clinically anxious and 14 control individuals before, during and after a stressful working memory task. Results Results indicate no between-group differences in HRV and HR at baseline. When starting the working memory task, the control group decreased significantly in HRV and the anxious group did not differ substantially in their change pattern from baseline to the start of the stressor. Finally, during the recovery phase of the working memory task, the clinically anxious and control individuals did not differ in their HFV or HR response compared to baseline. Conclusions From a clinical perspective, the results suggest that screening for the presence of anxiety disorders may help to identify patients with impaired HRV and HR functioning and to intervene on these important patient characteristics early in the treatment process.

Download Full-text

Interpreting neural decoding models using grouped model reliance

10.1101/656975 ◽

2019 ◽

Author(s):

Simon Valentin ◽

Maximilian Harkotte ◽

Tzvetan Popov

Keyword(s):

Machine Learning ◽

Working Memory ◽

Memory Load ◽

Memory Task ◽

Machine Learning Algorithms ◽

Support Vector ◽

Neural Decoding ◽

Working Memory Load ◽

Eeg Data

AbstractThe application of machine learning algorithms for decoding psychological constructs based on neural data is becoming increasingly popular. However, there is a need for methods that allow to interpret trained decoding models, as a step towards bridging the gap between theory-driven cognitive neuroscience and data-driven decoding approaches. The present study demonstrates grouped model reliance as a model-agnostic permutation-based approach to this problem. Grouped model reliance indicates the extent to which a trained model relies on conceptually related groups of variables, such as frequency bands or regions of interest in electroencephalographic (EEG) data. As a case study to demonstrate the method, random forest and support vector machine models were trained on within-participant single-trial EEG data from a Sternberg working memory task. Participants were asked to memorize a sequence of digits (0–9), varying randomly in length between one, four and seven digits, where EEG recordings for working memory load estimation were taken from a 3-second retention interval. Present results confirm previous findings in so far, as both random forest and support vector machine models relied on alpha-band activity in most subjects. However, as revealed by further analyses, patterns in frequency and particularly topography varied considerably between individuals, pointing to more pronounced inter-individual differences than reported previously.Author summaryModern machine learning algorithms currently receive considerable attention for their predictive power in neural decoding applications. However, there is a need for methods that make such predictive models interpretable. In the present work, we address the problem of assessing which aspects of the input data a trained model relies upon to make predictions. We demonstrate the use of grouped model-reliance as a generally applicable method for interpreting neural decoding models. Illustrating the method on a case study, we employed an experimental design in which a comparably small number of participants (10) completed a large number of trials (972) over multiple electroencephalography (EEG) recording sessions from a Sternberg working memory task. Trained decoding models consistently relied on alpha frequency activity, which is in line with existing research on the relationship between neural oscillations and working memory. However, our analyses also indicate large inter-individual variability with respect to the relation between activity patterns and working memory load in frequency and topography. Taken together, we argue that grouped model reliance provides a useful tool to better understand the workings of (sometimes otherwise black-box) decoding models.

Download Full-text

Event-related components are structurally represented by intrinsic event-related potentials

Scientific Reports ◽

10.1038/s41598-021-85235-0 ◽

2021 ◽

Vol 11 (1) ◽

Author(s):

Chong-Chih Tsai ◽

Wei-Kuang Liang

Keyword(s):

Cognitive Task ◽

Memory Task ◽

Event Related Potentials ◽

Mode Analysis ◽

Data Sets ◽

Brain Functions ◽

Eeg Data ◽

Related Potentials ◽

Statistical Effects ◽

Filtering Procedure

AbstractThe detection of event-related potentials (ERPs) through electroencephalogram (EEG) analysis is a well-established method for understanding brain functions during a cognitive process. To increase the signal-to-noise ratio (SNR) and stationarity of the data, ERPs are often filtered to a wideband frequency range, such as 0.05–30 Hz. Alternatively, a natural-filtering procedure can be performed through empirical mode decomposition (EMD), which yields intrinsic mode functions (IMFs) for each trial of the EEG data, followed by averaging over trials to generate the event-related modes. However, although the EMD-based filtering procedure has advantages such as a high SNR, suitable waveform shape, and high statistical power, one fundamental drawback of the procedure is that it requires the selection of an IMF (or a partial sum of a range of IMFs) to determine an ERP component effectively. Therefore, in this study, we propose an intrinsic ERP (iERP) method to overcome the drawbacks and retain the advantages of event-related mode analysis for investigating ERP components. The iERP method can reveal multiple ERP components at their characteristic time scales and suitably cluster statistical effects among modes by using a tailored definition of each mode’s neighbors. We validated the iERP method by using realistic EEG data sets acquired from a face perception task and visual working memory task. By using these two data sets, we demonstrated how to apply the iERP method to a cognitive task and incorporate existing cluster-based tests into iERP analysis. Moreover, iERP analysis revealed the statistical effects between (or among) experimental conditions more effectively than the conventional ERP method did.

Download Full-text

Increased cognitive workload evokes greater neurovascular coupling responses in healthy young adults

PLoS ONE ◽

10.1371/journal.pone.0250043 ◽

2021 ◽

Vol 16 (5) ◽

pp. e0250043

Author(s):

Tamas Csipo ◽

Agnes Lipecz ◽

Peter Mukli ◽

Dhay Bahadli ◽

Osamah Abdulhussein ◽

...

Keyword(s):

Working Memory ◽

Cognitive Impairment ◽

Prefrontal Cortex ◽

Cognitive Task ◽

Transcranial Doppler Sonography ◽

Memory Task ◽

Neurovascular Coupling ◽

Cognitive Stimulation ◽

Cognitive Test ◽

Functional Near Infrared Spectroscopy

Understanding how the brain allocates resources to match the demands of active neurons under physiological conditions is critically important. Increased metabolic demands of active brain regions are matched with hemodynamic responses known as neurovascular coupling (NVC). Several methods that allow noninvasive assessment of brain activity in humans detect NVC and early detection of NVC impairment may serve as an early marker of cognitive impairment. Therefore, non-invasive NVC assessments may serve as a valuable tool to detect early signs of cognitive impairment and dementia. Working memory tasks are routinely employed in the evaluation of cognitive task-evoked NVC responses. However, recent attempts that utilized functional near-infrared spectroscopy (fNIRS) or transcranial Doppler sonography (TCD) while using a similar working memory paradigm did not provide convincing evidence for the correlation of the hemodynamic variables measured by these two methods. In the current study, we aimed to compare fNIRS and TCD in their performance of differentiating NVC responses evoked by different levels of working memory workload during the same working memory task used as cognitive stimulation. Fourteen healthy young individuals were recruited for this study and performed an n-back cognitive test during TCD and fNIRS monitoring. During TCD monitoring, the middle cerebral artery (MCA) flow was bilaterally increased during the task associated with greater cognitive effort. fNIRS also detected significantly increased activation during a more challenging task in the left dorsolateral prefrontal cortex (DLPFC), and in addition, widespread activation of the medial prefrontal cortex (mPFC) was also revealed. Robust changes in prefrontal cortex hemodynamics may explain the profound change in MCA blood flow during the same cognitive task. Overall, our data support our hypothesis that both TCD and fNIRS methods can discriminate NVC evoked by higher demand tasks compared to baseline or lower demand tasks.

Download Full-text