A very common strategy for rejecting electroencephalographic (EEG) artifacts, includes the decomposition of filtered EEG signals using a Blind Source Separation (BSS) algorithm, the identification and removal of artifactual components and the reconstruction of the cleaned EEG signals. In this pipeline, the performance of the BSS algorithm, which is defined as its ability to separate properly the independent sources (like the EEG from artifactual sources), is very crucial for rejecting most of the artifacts, while maintaining the most part of EEG intact. The overwhelming majority of the published papers uses the extended INFOMAX version of Independent Component Analysis (ICA) for artifact rejection purposes. But is this the most efficient algorithm to separate EEG signals into independent components? This study comes to shed light to the aforementioned question by assessing the performance of the five most common BSS algorithms. The normalized entropy of the brain-related components, their correlation between independent components with the original sources and the amount of the overall mutual information reduction (MIR) achieved by each decomposition were computed in datasets with systematically varying numbers of electrodes (ranging from 19 tο 99), from 26 real human scalp EEG recordings. Additionally, 54 different datasets containing artificially contaminated EEG signals were also examined for the same purpose, on the basis of the Euclidean distance and the correlation, between the generated Independent Components (ICs) and the original vertical and horizontal eye signals, which were used for the contamination. The results support that the Adaptive Mixture ICA was the best performing BSS method.

Personality is the characteristic set of an individual’s behavioral and emotional patterns that evolve from biological and environmental factors. The recognition of personality profiles is crucial in making human–computer interaction (HCI) applications realistic, more focused, and user friendly. The ability to recognize personality using neuroscientific data underpins the neurobiological basis of personality. This paper aims to automatically recognize personality, combining scalp electroencephalogram (EEG) and machine learning techniques. As the resting state EEG has not so far been proven efficient for predicting personality, we used EEG recordings elicited during emotion processing. This study was based on data from the AMIGOS dataset reflecting the response of 37 healthy participants. Brain networks and graph theoretical parameters were extracted from cleaned EEG signals, while each trait score was dichotomized into low- and high-level using the k-means algorithm. A feature selection algorithm was used afterwards to reduce the feature-set size to the best 10 features to describe each trait separately. Support vector machines (SVM) were finally employed to classify each instance. Our method achieved a classification accuracy of 83.8% for extraversion, 86.5% for agreeableness, 83.8% for conscientiousness, 83.8% for neuroticism, and 73% for openness.

The need for automatic detection and classification of high-frequency oscillations (HFOs) as biomarkers of the epileptogenic tissue is strongly felt in the clinical field. In this context, the employment of artificial intelligence methods could be the missing piece to achieve this goal. This work proposed a double-step procedure based on machine learning algorithms and tested it on an intracranial electroencephalogram (iEEG) dataset available online. The first step aimed to define the optimal length for signal segmentation, allowing for an optimal discrimination of segments with HFO relative to those without. In this case, binary classifiers have been tested on a set of energy features. The second step aimed to classify these segments into ripples, fast ripples and fast ripples occurring during ripples. Results suggest that LDA applied to 10 ms segmentation could provide the highest sensitivity (0.874) and 0.776 specificity for the discrimination of HFOs from no-HFO segments. Regarding the three-class classification, non-linear methods provided the highest values (around 90%) in terms of specificity and sensitivity, significantly different to the other three employed algorithms. Therefore, this machine-learning-based procedure could help clinicians to automatically reduce the quantity of irrelevant data.

Mathematical anxiety (MA) is defined as a feeling of tension, apprehension, or fear that interferes with mathematical performance in various daily or academic situations. Cognitive consequences of MA have been studied a lot and revealed that MA seriously affects solving the complex problem due to the corruption of working memory (WM). The corruption of WM caused by MA is well documented in behavioral level, but the involved neurophysiological processes have not been properly addressed, despite the recent attention drawn on the neural basis of MA. This is the second part of our study that intents to investigate the neurophysiological aspects of MA and its implications to WM. In the first study, we saw how MA affects the early stages of numeric stimuli processes as the WM indirectly using event-related potentials in scalp electroencephalographic (EEG) signals. This paper goes one step further to investigate the cortical activations, obtained by the multichannel EEG recordings as well as the cortical functional networks in three WM tasks with increasing difficulty. Our results indicate that the high-math anxious (HMA) group activated more areas linked with negative emotions, pain, and fear, while the low-math anxious (LMA) group activated regions related to the encoding and retrieval processes of the WM. Functional connectivity analysis also reveals that the LMAs’ brain has got more structured cortical networks with increased connectivity in areas related to WM, such as the frontal cortex, while the HMAs’ brain has a more diffused and unstructured network, superimposing the evidence that the structured processes of WM are corrupted.