Search

Disorders of consciousness (DOC) are often related to serious changes in sleep structure. This article presents a sleep evaluation algorithm that scores the sleep structure of DOC patients to assist in assessing their consciousness level. The sleep evaluation algorithm is divided into two parts: 1) automatic sleep staging model: convolutional neural networks (CNNs) are employed for the extraction of signal features from electroencephalogram (EEG) and electrooculogram (EOG), and bidirectional long short-term memory (Bi-LSTM) with attention mechanism is applied to learn sequential information; and 2) consciousness assessment: automated sleep staging results are used to extract consciousness-related sleep features that are utilized by a support vector machine (SVM) classifier to assess consciousness. In this study, the CNN-BiLSTM model with an attention sleep network (CBASleepNet) was conducted using the sleep-EDF and MASS datasets. The experimental results demonstrated the effectiveness of the proposed model, which outperformed similar models. Moreover, CBASleepNet was applied to sleep staging in DOC patients through transfer learning and fine-tuning. Consciousness assessments were conducted on seven minimally conscious state (MCS) patients and four vegetative state (VS)/unresponsive wakefulness syndrome (UWS) patients, achieving an overall accuracy of 81.8%. The sleep evaluation algorithm can be used to evaluate the consciousness level of patients effectively.

An increasing number of people fail to properly regulate their emotions for various reasons. Although brain–computer interfaces (BCIs) have shown potential in neural regulation, few effective BCI systems have been developed to assist users in emotion regulation. In this article, we propose an electroencephalography (EEG)-based BCI for emotion regulation with virtual reality (VR) neurofeedback. Specifically, music clips with positive, neutral, and negative emotions were first presented, based on which the participants were asked to regulate their emotions. The BCI system simultaneously collected the participants’ EEG signals and then assessed their emotions. Furthermore, based on the emotion recognition results, the neurofeedback was provided to participants in the form of a facial expression of a virtual pop star on a three-dimensional (3-D) virtual stage. Eighteen healthy participants achieved satisfactory performance with an average accuracy of 81.1% with neurofeedback. Additionally, the average accuracy increased significantly from 65.4% at the start to 87.6% at the end of a regulation trial (a trial corresponded to a music clip). In comparison, these participants could not significantly improve the accuracy within a regulation trial without neurofeedback. The results demonstrated the effectiveness of our system and showed that VR neurofeedback played a key role during emotion regulation.

Despite previous efforts in depression detection studies, there is a scarcity of research on automatic depression detection using sleep structure, and several challenges remain: 1) how to apply sleep staging to detect depression and distinguish easily misjudged classes; and 2) how to adaptively capture attentive channel-dimensional information to enhance the interpretability of sleep staging methods. To address these challenges, an automatic sleep staging method based on a channel-temporal attention mechanism and a depression detection method based on sleep structure features are proposed. In sleep staging, a temporal attention mechanism is adopted to update the feature matrix, confidence scores are estimated for each sleep stage, the weight of each channel is adjusted based on these scores, and the final results are obtained through a temporal convolutional network. In depression detection, seven sleep structure features based on the results of sleep staging are extracted for depression detection between unipolar depressive disorder (UDD) patients, bipolar disorder (BD) patients, and healthy subjects. Experiments demonstrate the effectiveness of the proposed approaches, and the visualization of the channel attention mechanism illustrates the interpretability of our method. Additionally, this is the first attempt to employ sleep structure features to automatically detect UDD and BD in patients.

Objective: Brain-computer interfaces (BCIs) have tremendous application potential in communication, mechatronic control and rehabilitation. However, existing BCI systems are bulky, expensive and require laborious preparation before use. This study proposes a practical and user-friendly BCI system without compromising performance. Methods: A hybrid asynchronous BCI system was developed based on an elaborately designed wearable electroencephalography (EEG) amplifier that is compact, easy to use and offers a high signal-to-noise ratio (SNR). The wearable BCI system can detect P300 signals by processing EEG signals from three channels and operates asynchronously by integrating blink detection. Result: The wearable EEG amplifier obtains high quality EEG signals and introduces preprocessing capabilities to BCI systems. The wearable BCI system achieves an average accuracy of 94.03±4.65%, an average information transfer rate (ITR) of 31.42±7.39 bits/min and an average false-positive rate (FPR) of 1.78%. Conclusion: The experimental results demonstrate the feasibility and practicality of the developed wearable EEG amplifier and BCI system. Significance: Wearable asynchronous BCI systems with fewer channels are possible, indicating that BCI applications can be transferred from the laboratory to real-world scenarios.

In this paper, a novel spatio-temporal self-constructing graph neural network (ST-SCGNN) is proposed for cross-subject emotion recognition and consciousness detection. For spatio-temporal feature generation, activation and connection pattern features are first extracted and then combined to leverage their complementary emotion-related information. Next, a self-constructing graph neural network with a spatio-temporal model is presented. Specifically, the graph structure of the neural network is dynamically updated by the self-constructing module of the input signal. Experiments based on the SEED and SEED-IV datasets showed that the model achieved average accuracies of 85.90% and 76.37%, respectively. Both values exceed the state-of-the-art metrics with the same protocol. In clinical besides, patients with disorders of consciousness (DOC) suffer severe brain injuries, and sufficient training data for EEG-based emotion recognition cannot be collected. Our proposed ST-SCGNN method for cross-subject emotion recognition was first attempted in training in ten healthy subjects and testing in eight patients with DOC. We found that two patients obtained accuracies significantly higher than chance level and showed similar neural patterns with healthy subjects. Covert consciousness and emotion-related abilities were thus demonstrated in these two patients. Our proposed ST-SCGNN for cross-subject emotion recognition could be a promising tool for consciousness detection in DOC patients.

Electromagnetic source imaging (ESI) requires solving a highly ill-posed inverse problem. To seek a unique solution, traditional ESI methods impose various forms of priors that may not accurately reflect the actual source properties, which may hinder their broad applications. To overcome this limitation, in this article, a novel data-synthesized spatiotemporally convolutional encoder–decoder network (DST-CedNet) method is proposed for ESI. The DST-CedNet recasts ESI as a machine learning problem, where discriminative learning and latent-space representations are integrated in a CedNet to learn a robust mapping from the measured electroencephalography/magnetoencephalography (E/MEG) signals to the brain activity. In particular, by incorporating prior knowledge regarding dynamical brain activities, a novel data synthesis strategy is devised to generate large-scale samples for effectively training CedNet. This stands in contrast to traditional ESI methods where the prior information is often enforced via constraints primarily aimed for mathematical convenience. Extensive numerical experiments as well as analysis of a real MEG and epilepsy EEG dataset demonstrate that the DST-CedNet outperforms several state-of-the-art ESI methods in robustly estimating source signals under a variety of source configurations.

Network pruning has been widely studied to reduce the complexity of deep neural networks (DNNs) and hence speed up their inference. Unfortunately, most existing pruning methods ignore the changes in the model’s robustness before and after pruning, which makes pruned models vulnerable under dynamically perturbed environments (e.g., autonomous driving). Only a few works have explored the robustness of pruned models against adversarial attacks that significantly differ from perturbations in real-world scenarios. To bridge the gap between real-world applications and existing studies, in this work, we propose an adversarial pruning scheme, which automatically identifies and preserves robust channels to obtain robust pruned models that are suitable for practical deployment in dynamically perturbed environments. Specifically, to simulate real-world perturbations, we first employ multi-type adversarial attack samples and adversarial perturbation samples generated by an adversarial perturbation generator to create mixed noise samples. Then, we propose a plug-and-play feature scoring module and a novel contribution difference loss to evaluate the robustness of intermediate features dynamically. Next, to leverage robust intermediate features to identify robust channels, we have developed a simple but effective gating mechanism that evaluates the robustness of channels and preserves robust channels during training. Lastly, we compress the model in a layer-wise or block-wise manner. Compared to existing methods, our scheme enhances the robustness of the pruned model in a broader sense, making it better able to against dynamic perturbations in the real world. Extensive experimental results on well-known dataset benchmarks and popular network architectures demonstrate the effectiveness of our method.

Acute myeloid leukaemia (AML) is a malignant disease of myeloid hematopoietic stem/progenitor cells. Despite improved understanding of the pathogenesis of AML since the 1980s, the standard treatment for AML has remained virtually unchanged. Numerous studies have found poor survival rates and high relapse rates among older patients with AML. Several novel therapies for AML, including cytotoxic drugs, genetically- and epigenetically-targeted drugs, and immunotherapies, have been developed in recent years. Alternative treatments with improved efficacy are required for AML because many patients cannot tolerate the toxic effects of chemotherapy. Non-coding RNAs, including microRNAs (miRNAs), long noncoding RNAs (lncRNAs) and circular RNAs (circRNAs), are attractive treatment targets for cancers and several other diseases. LncRNAs, miRNAs and circRNAs regulate DNA transcription and translation. Over the past decade, significant efforts have been made to develop RNA-based therapies, mainly antisense oligonucleotides and small interfering RNA, some of which have been approved for clinical use. Here we reviewed the mechanisms underlying the in vitroand in vivoeffects of promising targets and potential drugs, focusing on the drugs most likely to be used for clinical treatment, to aid in the development of precision therapy for AML.