International Journal of Neural Systems

We investigated the functional network reorganization caused by low-frequency electrical stimulation (LFES) of human brain cortical surface. Intracranial EEG data from subdural grid positions were analyzed in 16 pre-surgery epileptic patients. LFES was performed by injecting current pulses (10mA, 0.2ms pulse width, 0.5Hz, 25 trials) into all adjacent electrode contacts. Dynamic functional connectivity analysis was carried out on two frequency bands (low: 1–4Hz; high: 10–40Hz) to investigate the early, high frequency and late, low frequency responses elicited by the stimulation. The centralization increased in early compared to late responses, suggesting a more prominent role of direct neural links between primarily activated areas and distant brain regions. Injecting the current into the seizure onset zone (SOZ) evoked a more integrated functional topology during the early (N1) period of the response, whereas during the late (N2) period — regardless of the stimulation site — the connectedness of the SOZ was elevated compared to the non-SOZ tissue. The abnormal behavior of the epileptic sub-network during both part of the responses supports the idea of the pathogenic role of impaired inhibition and excitation mechanisms in epilepsy.

We evaluated the properties of oscillations in the Mayer waves (MW) frequency range ($\sim 0.1$Hz) detected in blood pressure, heart rate variability, cerebral blood oxygenation changes and evolution of electroencephalographic (EEG) rhythms to elucidate the mechanisms of MW generation. We examined the persistence of MW in different signals and stability of their oscillations on the level of individual MW waveforms, which was achieved by applying matching pursuit (MP). MP yields adaptive time-frequency approximation of signal’s structures in terms of frequency, amplitude, time occurrence, and time-span. The number of waveforms contributing to 95% of the energy of the signals was vastly different for the time series, but the average number of waveforms conforming to the MW criteria was almost the same ($3.5\pm 0.4$ per 120s epoch). In all the investigated signals, MW had the same distributions of frequency and the number of cycles. We show that the MW energy ratios in different signals varied strongly, $p<0.001$. The highest percentage of MW energy was observed in blood pressure signals, heart rate variability, and reduced hemoglobin, in contrast to brain signals and oxygenated hemoglobin. The percentage of MW energy was related to the strength of causal influence exerted by them on the other signals. Our results indicate existence of a common mechanism of MW generation and support the hypothesis of MW generation in the baroreflex loop; however, they do not exclude the action of a central pacemaker.

Recent continuous task studies, such as narrative speech comprehension, show that fluctuations in brain functional connectivity (FC) are altered and enhanced compared to the resting state. Here, we characterized the fluctuations in FC during comprehension of speech and time-reversed speech conditions. The correlations of Hilbert envelope of source-level EEG data were used to quantify FC between spatially separate brain regions. A symmetric multivariate leakage correction was applied to address the signal leakage issue before calculating FC. The dynamic FC was estimated based on a sliding time window. Then, principal component analysis (PCA) was performed on individually concatenated and temporally concatenated FC matrices to identify FC patterns. We observed that the mode of FC induced by speech comprehension can be characterized with a single principal component. The condition-specific FC demonstrated decreased correlations between frontal and parietal brain regions and increased correlations between frontal and temporal brain regions. The fluctuations of the condition-specific FC characterized by a shorter time demonstrated that dynamic FC also exhibited condition specificity over time. The FC is dynamically reorganized and FC dynamic pattern varies along a single mode of variation during speech comprehension. The proposed analysis framework seems valuable for studying the reorganization of brain networks during continuous task experiments.

Despite many advances in the development of retinal prostheses, clinical reports show that current retinal prosthesis subjects can only perceive prosthetic vision with poor visual acuity. A possible approach for improving visual acuity is to produce virtual electrodes (VEs) through electric field modulation. Generating controllable and localized VEs is a crucial factor in effectively improving the perceptive resolution of the retinal prostheses. In this paper, we aimed to design a microelectrode array (MEA) that can produce converged and controllable VEs by current steering stimulation strategies. Through computational modeling, we designed a three-dimensional concentric ring–disc MEA and evaluated its performance with different stimulation strategies. Our simulation results showed that electrode–retina distance (ERD) and inter-electrode distance (IED) can dramatically affect the distribution of electric field. Also the converged VEs could be produced when the parameters of the three-dimensional MEA were appropriately set. VE sites can be controlled by manipulating the proportion of current on each adjacent electrode in a current steering group (CSG). In addition, spatial localization of electrical stimulation can be greatly improved under quasi-monopolar (QMP) stimulation. This study may provide support for future application of VEs in epiretinal prosthesis for potentially increasing the visual acuity of prosthetic vision.

Olfactory-induced electroencephalogram (EEG) signal classification is of great significance in a variety of fields, such as disorder treatment, neuroscience research, multimedia applications and brain–computer interface. In this paper, a trapezoid difference-based electrode sequence hashing method is proposed for olfactory EEG signal classification. First, an $N$-layer trapezoid feature set whose size ratio of the top, bottom and height is 1:2:1 is constructed for each frequency band of each EEG sample. This construction is based on $N$ optimized power-spectral-density features extracted from $N$ real electrodes and $N$ nonreal electrode’s features. Subsequently, the $N$ real electrodes’ sequence (ES) codes of each layer of the constructed trapezoid feature set are obtained by arranging the feature values in ascending order. Finally, the nearest neighbor classification is used to find a class whose ES codes are the most similar to those of the testing sample. Thirteen-class olfactory EEG signals collected from 11 subjects are used to compare the classification performance of the proposed method with six traditional classification methods. The comparison shows that the proposed method gives average accuracy of 94.3%, Cohen’s kappa value of 0.94, precision of 95.0%, and F1-measure of 94.6%, which are higher than those of the existing methods.

Traditional training methods need to collect a large amount of data for every subject to train a subject-specific classifier, which causes subjects fatigue and training burden. This study proposes a novel training method, TrAdaBoost based on cross-validation and an adaptive threshold (CV-T-TAB), to reduce the amount of data required for training by selecting and combining multiple subjects’ classifiers that perform well on a new subject to train a classifier. This method adopts cross-validation to extend the amount of the new subject’s training data and sets an adaptive threshold to select the optimal combination of the classifiers. Twenty-five subjects participated in the N200- and P300-based brain–computer interface. The study compares CV-T-TAB to five traditional training methods by testing them on the training of a support vector machine. The accuracy, information transfer rate, area under the curve, recall and precision are used to evaluate the performances under nine conditions with different amounts of data. CV-T-TAB outperforms the other methods and retains a high accuracy even when the amount of data is reduced to one-third of the original amount. The results imply that CV-T-TAB is effective in improving the performance of a subject-specific classifier with a small amount of data by adopting multiple subjects’ classifiers, which reduces the training cost.