Search

In clinical practice, algorithmic predictions may seriously jeopardise patients' health and thus are required to be validated by medical experts before a final clinical decision is met. Towards that aim, there is need to incorporate explainable artificial intelligence techniques into medical research. In the specific field of epileptic seizure detection there are several machine learning algorithms but less methods on explaining them in an interpretable way. Therefore, we introduce XAI4EEG: an application-aware approach for an explainable and hybrid deep learning-based detection of seizures in multivariate EEG time series. In XAI4EEG, we combine deep learning models and domain knowledge on seizure detection, namely (a) frequency bands, (b) location of EEG leads and (c) temporal characteristics. XAI4EEG encompasses EEG data preparation, two deep learning models and our proposed explanation module visualizing feature contributions that are obtained by two SHAP explainers, each explaining the predictions of one of the two models. The resulting visual explanations provide an intuitive identification of decision-relevant regions in the spectral, spatial and temporal EEG dimensions. To evaluate XAI4EEG, we conducted a user study, where users were asked to assess the outputs of XAI4EEG, while working under time constraints, in order to emulate the fact that clinical diagnosis is done - more often than not - under time pressure. We found that the visualizations of our explanation module (1) lead to a substantially lower time for validating the predictions and (2) leverage an increase in interpretability, trust and confidence compared to selected SHAP feature contribution plots. [ABSTRACT FROM AUTHOR]

We here study a network of synaptic relations mingling excitatory and inhibitory neuron nodes that displays oscillations quite similar to electroencephalogram (EEG) brain waves, and identify abrupt variations brought about by swift synaptic mediations. We thus conclude that corresponding changes in EEG series surely come from the slowdown of the activity in neuron populations due to synaptic restrictions. The latter happens to generate an imbalance between excitation and inhibition causing a quick explosive increase of excitatory activity, which turns out to be a (first-order) transition among dynamic mental phases. Moreover, near this phase transition, our model system exhibits waves with a strong component in the so-called delta-theta domain that coexist with fast oscillations. These findings provide a simple explanation for the observed delta-gamma and theta-gamma modulation in actual brains, and open a serious and versatile path to understand deeply large amounts of apparently erratic, easily accessible brain data.

Graphical analysis of complex brain networks is a fundamental area of modern neuroscience. Functional connectivity is important since many neurological and psychiatric disorders, including schizophrenia, are described as 'dys-connectivity' syndromes. Using electroencephalogram time series collected on each of a group of 15 individuals with a common medical diagnosis of positive syndrome schizophrenia we seek to build a single, representative, brain functional connectivity group graph. Disparity/distance measures between spectral matrices are identified and used to define the normalized graph Laplacian enabling clustering of the spectral matrices for detecting 'outlying' individuals. Two such individuals are identified. For each remaining individual, we derive a test for each edge in the connectivity graph based on average estimated partial coherence over frequencies, and associated p-values are found. For each edge these are used in a multiple hypothesis test across individuals and the proportion rejecting the hypothesis of no edge is used to construct a connectivity group graph. This study provides a framework for integrating results on multiple individuals into a single overall connectivity structure. [ABSTRACT FROM AUTHOR]

The aim of the present research was to measure the brain order/chaos inter-hemispheric asymmetry, long memory processing, and predictability of an EEG signal at different timescales during a simple exercise of closed eyes imagination or mental projection. We studied inter-hemispheric asymmetries in frontal, temporal and occipital areas of the brain during 2 minutes of silent and relaxed mental projective task of visualizing themselves performing a physical exercise practice. EEG signal was recorded using the brain-computer device EmotivEpoc® from 8 subjects aged between 20 and 30 years, all fellowships in two weeks of modern dance workshop. The results showed that Hurst (H) exponent used as anestimator of the persistence/anti-persistence dynamics of the EEG process vary inversely with the length of the time window (1, 3, 12, 30 and 120 s) used to calculate H at different timescales.We foundsignificant inter-hemispheric asymmetry in the temporal and occipital regions, with left hemisphere having the higher scores of H. Together with a common general trend we observed inter-individual variability of inter-hemispheric asymmetry between frontal, temporal and occipital lobes, being the temporal the one with the highest scores. We evaluated statistically these differences and significance. [ABSTRACT FROM AUTHOR]

Simple Summary: In this study, we explore the emergence of oscillatory behavior similar to the signals of brain activity observed in electroencephalograms (EEGs) using a network of synaptic relations mingling excitatory and inhibitory neuron nodes. We identify abrupt variations on that activity brought about by swift synaptic mediations. These changes are originated by the slowdown of the activity of inhibitory neuron populations due to synaptic depression. The latter generates an imbalance between excitation and inhibition causing a quick explosive increase of excitatory activity, which turns out to be a (first-order) phase transition among different oscillatory states. Interestingly enough, near this transition, our model system exhibits oscillatory activity with a strong component in the delta-theta domain that coexist with fast oscillations and happens to be similar to the observed delta-gamma and theta-gamma modulation in actual brains. Our findings here help to understand actual brain activity data in terms of nonequilibrium phase transitions theory. We here study a network of synaptic relations mingling excitatory and inhibitory neuron nodes that displays oscillations quite similar to electroencephalogram (EEG) brain waves, and identify abrupt variations brought about by swift synaptic mediations. We thus conclude that corresponding changes in EEG series surely come from the slowdown of the activity in neuron populations due to synaptic restrictions. The latter happens to generate an imbalance between excitation and inhibition causing a quick explosive increase of excitatory activity, which turns out to be a (first-order) transition among dynamic mental phases. Moreover, near this phase transition, our model system exhibits waves with a strong component in the so-called delta-theta domain that coexist with fast oscillations. These findings provide a simple explanation for the observed delta-gamma and theta-gamma modulation in actual brains, and open a serious and versatile path to understand deeply large amounts of apparently erratic, easily accessible brain data. [ABSTRACT FROM AUTHOR]