A Multibranch of Convolutional Neural Network Models for Electroencephalogram-Based Motor Imagery Classification

Automatic high-level feature extraction has become a possibility with the advancement of deep learning, and it has been used to optimize efficiency. Recently, classification methods for convolutional neural network (CNN)-based electroencephalography (EEG) motor imagery have been proposed, and have achieved reasonably high classification accuracy. These approaches, however, use the CNN single convolution scale, whereas the best convolution scale varies from subject to subject. This limits the precision of classification. This paper proposes multibranch CNN models to address this issue by effectively extracting the spatial and temporal features from raw EEG data, where the branches correspond to different filter kernel sizes. The proposed method’s promising performance is demonstrated by experimental results on two public datasets, the BCI Competition IV 2a dataset and the High Gamma Dataset (HGD). The results of the technique show a 9.61% improvement in the classification accuracy of multibranch EEGNet (MBEEGNet) from the fixed one-branch EEGNet model, and 2.95% from the variable EEGNet model. In addition, the multibranch ShallowConvNet (MBShallowConvNet) improved the accuracy of a single-scale network by 6.84%. The proposed models outperformed other state-of-the-art EEG motor imagery classification methods.

Localized Task-Invariant Emotional Valence Encoding Revealed by Intracranial Recordings

The ability to distinguish between negative, positive and neutral valence is a key part of emotion perception. Emotional valence has conceptual meaning that supersedes any particular type of stimulus, although it is typically captured experimentally in association with particular tasks. We sought to identify neural encoding for task-invariant emotional valence. We evaluated whether high gamma responses (HGRs) to visually displayed words conveying emotions could be used to decode emotional valence from HGRs to facial expressions. Intracranial electroencephalography (iEEG) was recorded from fourteen individuals while they participated in two tasks, one involving reading words with positive, negative, and neutral valence, and the other involving viewing faces with positive, negative, and neutral facial expressions. Quadratic discriminant analysis was used to identify information in the HGR that differentiates the three emotion conditions. A classifier was trained on the emotional valence labels from one task and was cross-validated on data from the same task (within-task classifier) as well as the other task (between-task classifier). Emotional valence could be decoded in the left medial orbitofrontal cortex and middle temporal gyrus, both using within-task classifiers as well as between-task classifiers. These observations suggest the presence of task-independent emotional valence information in the signals from these regions.

Multiple gamma rhythms carry distinct spatial frequency information in primary visual cortex

Gamma rhythms in many brain regions, including the primary visual cortex (V1), are thought to play a role in information processing. Here, we report a surprising finding of 3 narrowband gamma rhythms in V1 that processed distinct spatial frequency (SF) signals and had different neural origins. The low gamma (LG; 25 to 40 Hz) rhythm was generated at the V1 superficial layer and preferred a higher SF compared with spike activity, whereas both the medium gamma (MG; 40 to 65 Hz), generated at the cortical level, and the high gamma HG; (65 to 85 Hz), originated precortically, preferred lower SF information. Furthermore, compared with the rates of spike activity, the powers of the 3 gammas had better performance in discriminating the edge and surface of simple objects. These findings suggest that gamma rhythms reflect the neural dynamics of neural circuitries that process different SF information in the visual system, which may be crucial for multiplexing SF information and synchronizing different features of an object.

Decreased Phase–Amplitude Coupling Between the mPFC and BLA During Exploratory Behaviour in Chronic Unpredictable Mild Stress-Induced Depression Model of Rats

Depression is a common neuropsychiatric illness observed worldwide, and reduced interest in exploration is one of its symptoms. The control of dysregulated medial prefrontal cortex (mPFC) over the basolateral amygdala (BLA) is related to depression. However, the oscillation interaction in the mPFC-BLA circuit has remained elusive. Therefore, this study used phase–amplitude coupling (PAC), which provides complicated forms of information transmission by the phase of low-frequency rhythm, modulating the amplitude of high-frequency rhythm, and has a potential application for the treatment of neurological disease. The chronic unpredictable mild stress (CUMS) was used to prepare the rat models of depression. Moreover, multichannel in vivo recording was applied to obtain the local field potentials (LFPs) of the mPFC, the BLA in rats in control, and CUMS groups, while they explored the open field. The results showed prominent coupling between the phase of theta oscillation (4–12 Hz) in the mPFC and the amplitude of high-gamma oscillation (70–120 Hz) in the BLA. Compared to the control group, this theta–gamma PAC was significantly decreased in the CUMS group, which was accompanied by the diminished exploratory behaviour. The results indicate that the coupling between the phase of theta in the mPFC and the amplitude of gamma in the BLA is involved in exploratory behaviour, and this decreased coupling may inhibit exploratory behaviour of rats exposed to CUMS.

Neural Oscillatory Characteristics of Feedback Associated Activity in Globus Pallidus Interna

Neural oscillatory activities in basal ganglia have prominent roles in cognitive processes on local and global scales. However, the characteristics of high frequency oscillatory activities during cognitive tasks have not been extensively explored in human Globus Pallidus internus (GPi). This study aimed to investigate amplitude and interhemispheric coupling of bilateral GPi high gamma bursts in dystonia and Parkinson's Disease (PD) patients, in on and off medication states, after feedback during the Intra-Extra-Dimension shift (IED) task. Bilateral GPi Local Field Potentials (LFP) activity was recorded via externalized DBS electrodes during the IED task. Inter hemisphere phase synchrony was assessed using Inter-Site Phase Clustering (ISPC). Transient high gamma activity (~100-150Hz) was observed immediately after feedback in the dystonia patient. Moreover, these bursts were phase synchronous between left and right GPis with an antiphase clustering of phase differences. In contrast, no synchronous high gamma activity was detected in the PD patient with or without dopamine administration. The off-med PD patient displayed enhanced low frequency clusters ameliorated by medication in the on-med state. Furthermore, an increased low frequency activity was observed after feedback of incorrect trials in both disease states. The current study provides a rare report of antiphase homotopic synchrony in human GPi, potentially related to incorporating and processing feedback information. The absence of these activities in off and on-med PD indicates the potential presence of impaired medication independent circuits related to feedback processing. Together, these findings are helpful in pointing to the potential role of GPi's synchronized high frequency activity in cognitive tasks and feedback information processing.

The hippocampus as the switchboard between perception and memory

Adaptive memory recall requires a rapid and flexible switch from external perceptual reminders to internal mnemonic representations. However, owing to the limited temporal or spatial resolution of brain imaging modalities used in isolation, the hippocampal–cortical dynamics supporting this process remain unknown. We thus employed an object-scene cued recall paradigm across two studies, including intracranial electroencephalography (iEEG) and high-density scalp EEG. First, a sustained increase in hippocampal high gamma power (55 to 110 Hz) emerged 500 ms after cue onset and distinguished successful vs. unsuccessful recall. This increase in gamma power for successful recall was followed by a decrease in hippocampal alpha power (8 to 12 Hz). Intriguingly, the hippocampal gamma power increase marked the moment at which extrahippocampal activation patterns shifted from perceptual cue toward mnemonic target representations. In parallel, source-localized EEG alpha power revealed that the recall signal progresses from hippocampus to posterior parietal cortex and then to medial prefrontal cortex. Together, these results identify the hippocampus as the switchboard between perception and memory and elucidate the ensuing hippocampal–cortical dynamics supporting the recall process.

Engineering Nonlinear Epileptic Biomarkers Using Deep Learning and Benford’s Law

In this study we designed two deep neural networks to encode 16 feature latent spaces for early seizure detection in intracranial EEG and compared them to 16 widely used engineered metrics: Epileptogenicity Index (EI), Phase Locked High Gamma (PLHG), Time and Frequency Domain Cho Gaines Distance (TDCG, FDCG), relative band powers, and log absolute band powers (from alpha, beta, theta, delta, gamma, and high gamma bands. The deep learning models were pretrained for seizure identification on the time and frequency domains of one second single channel clips of 127 seizures (from 25 different subjects) using “leave-one-out” (LOO) cross validation. Each neural network extracted unique feature spaces that were used to train a Random Forest Classifier (RFC) for seizure identification and latency tasks. The Gini Importance of each feature was calculated from the pretrained RFC, enabling the most significant features (MSFs) for each task to be identified. The MSFs were extracted from the UPenn and Mayo Clinic's Seizure Detection Challenge to train another RFC for the contest. They obtained an AUC score of 0.93, demonstrating a transferable method to identify interpretable biomarkers for seizure detection.

High-gamma mirror activity patterns in the human brain during reach-to-grasp movement observation, retention, and execution—An MEG study

While the existence of a human mirror neuron system is evident, the involved brain areas and their exact functional roles remain under scientific debate. A number of functionally different mirror neuron types, neurons that selectively respond to specific grasp phases and types for example, have been reported with single cell recordings in monkeys. In humans, spatially limited, intracranially recorded electrophysiological signals in the high-gamma (HG) range have been used to investigate the human mirror system, as they are associated with spiking activity in single neurons. Our goal here is to complement previous intracranial HG studies by using magnetoencephalography to record HG activity simultaneously from the whole head. Participants performed a natural reach-to-grasp movement observation and delayed imitation task with different everyday objects and grasp types. This allowed us to characterize the spatial organization of cortical areas that show HG-activation modulation during movement observation (mirroring), retention (mnemonic mirroring), and execution (motor control). Our results show mirroring related HG modulation patterns over bilateral occipito-parietal as well as sensorimotor areas. In addition, we found mnemonic mirroring related HG modulation over contra-lateral fronto-temporal areas. These results provide a foundation for further human mirror system research as well as possible target areas for brain-computer interface and neurorehabilitation approaches.

α-Synuclein aggregation in the olfactory bulb induces olfactory deficits by perturbing granule cells and granular–mitral synaptic transmission

Olfactory dysfunction is an early pre-motor symptom of Parkinson’s disease (PD) but the neural mechanisms underlying this dysfunction remain largely unknown. Aggregation of α-synuclein is observed in the olfactory bulb (OB) during the early stages of PD, indicating a relationship between α-synuclein pathology and hyposmia. Here we investigate whether and how α-synuclein aggregates modulate neural activity in the OB at the single-cell and synaptic levels. We induced α-synuclein aggregation specifically in the OB via overexpression of double-mutant human α-synuclein by an adeno-associated viral (AAV) vector. We found that α-synuclein aggregation in the OB decreased the ability of mice to detect odors and to perceive attractive odors. The spontaneous activity and odor-evoked firing rates of single mitral/tufted cells (M/Ts) were increased by α-synuclein aggregates with the amplitude of odor-evoked high-gamma oscillations increased. Furthermore, the decreased activity in granule cells (GCs) and impaired inhibitory synaptic function were responsible for the observed hyperactivity of M/Ts induced by α-synuclein aggregates. These results provide direct evidences of the role of α-synuclein aggregates on PD-related olfactory dysfunction and reveal the neural circuit mechanisms by which olfaction is modulated by α-synuclein pathology.

Basolateral amygdala to posterior piriform cortex connectivity ensures precision in learned odor threat

Odor perception can both evoke emotional states and be shaped by emotional or hedonic states. The amygdala complex plays an important role in recognition of, and response to, hedonically valenced stimuli, and has strong, reciprocal connectivity with the primary olfactory (piriform) cortex. Here, we used differential odor-threat conditioning in rats to test the role of basolateral amygdala (BLA) input to the piriform cortex in acquisition and expression of learned olfactory threat responses. Using local field potential recordings, we demonstrated that functional connectivity (high gamma band coherence) between the BLA and posterior piriform cortex (pPCX) is enhanced after differential threat conditioning. Optogenetic suppression of activity within the BLA prevents learned threat acquisition, as do lesions of the pPCX prior to threat conditioning (without inducing anosmia), suggesting that both regions are critical for acquisition of learned odor threat responses. However, optogenetic BLA suppression during testing did not impair threat response to the CS+ , but did induce generalization to the CS−. A similar loss of stimulus control and threat generalization was induced by selective optogenetic suppression of BLA input to pPCX. These results suggest an important role for amygdala-sensory cortical connectivity in shaping responses to threatening stimuli.