Numerous studies have made significant contributions to understanding resting-state brain networks, advancing the field of neuroscience. Studying dynamic functional connectivity is essential for capturing the temporal evolution of brain network organization. However, investigations of task-related functional networks, especially those assessed through dynamic connectivity during continuous cognitive tasks, remain relatively sparse. This study aims to investigate the temporal dynamics of functional prefrontal networks during continuous cognitive control through dynamic functional connectivity analysis. In contrast to conventional methods that primarily focus on identifying static spatial connectivity patterns, this study applies temporal group independent component analysis (TG-ICA) to functional near-infrared spectroscopy (fNIRS) data acquired during a continuous Stroop Color-Word task. This approach enables the identification of temporally evolving functional connectivity networks at the group level. The results revealed three distinct and interpretable prefrontal network components, including a left-lateralized system for early rule implementation and executive control, a right-dominant network for conflict monitoring and attentional reallocation, and a bilateral network supporting sustained goal maintenance and cognitive stability. These networks exhibited time-varying engagement aligned with different stages of cognitive control during continuous tasks. The findings highlight the utility of TG-ICA in capturing the spatiotemporal characteristics of functional brain networks and offer new insights into the dynamic organization of the prefrontal cortex during executive functioning.

Sleep research has evolved considerably since the first sleep electroencephalography recordings in the 1930s and the discovery of well-distinguishable sleep stages in the 1950s. While electrophysiological recordings have been used to describe the sleeping brain in much detail, since the 1990s neuroimaging techniques have been applied to uncover the brain organization and functional connectivity of human sleep with greater spatial resolution. The combination of electroencephalography with different neuroimaging modalities such as positron emission tomography, structural magnetic resonance imaging and functional magnetic resonance imaging imposes several challenges for sleep studies, for instance, the need to combine polysomnographic recordings to assess sleep stages accurately, difficulties maintaining and consolidating sleep in an unfamiliar and restricted environment, scanner-induced distortions with physiological artefacts that may contaminate polysomnography recordings, and the necessity to account for all physiological changes throughout the sleep cycles to ensure better data interpretability. Here, we review the field of sleep neuroimaging in healthy non-sleep-deprived populations, from early findings to more recent developments. Additionally, we discuss the challenges of applying concurrent electroencephalography and imaging techniques to sleep, which consequently have impacted the sample size and generalizability of studies, and possible future directions for the field.

The EMG-based neural control strategy provides technical support and innovative methodologies for the precise control and natural movement of hand prostheses. For enabling the prostheses to mimic the function of the human hand, it is essential to fully understand the principle of neuromuscular control of human hand. However, we still have little knowledge regarding the functional significance of the proportion of common or independent neural input to the antagonist muscles in extensor and flexor during performing different tasks. Therefore, we used a discharge coherence analysis of motor unit (MU) spike trains to investigate different sources of common and independent input between extensor and flexor muscle groups in different hand gestures. 14 gestures were selected, including the extension of individual and multi fingers, as well as pinch finger tasks. The MU spike trains were obtained from the decomposition of high-density surface electromyography (HD-sEMG) recordings using fastICA. The proportion of the independent neural input to extensor or flexor muscle groups was calculated by the ratio of residual coherence to total coherence. The results showed that the two muscle groups are comparable in independent proportion within each muscle group with relatively greater values in the extensor muscle. The degree of common input shared between the two muscle groups exhibited the highest level (around 35%) in the delta band (1-4Hz) compared to a very small proportion (<13%) in the other frequency bands. Additionally, a significant difference was observed in little finger extension and hand close tasks compared to other hand gestures. Overall, the varied proportions of independent neural input across hand gestures and among muscle groups illustrate the precise neural modulation involved in the co-contraction of the flexor and extensor muscles during flexible hand movements, as observed through the microscopic view of motoneurons.

Graph neural networks show strong capability of learning spatial relationships between channels. In recent studies, they greatly advanced automatic epileptic seizures detection via multi-channels scalp electroencephalography (EEG). In this work, we used Graph WaveNet to extract spatial and temporal dependencies of epileptic seizures. However, EEG signals often contain strong noise, leading to unsatisfactory model performance. This study compared effects of four input preprocessing strategies on model robustness. The fast Fourier transform (FFT) features, the input of the network, were preprocessed by intact, hard, learnable, and random selection. Results show that the Graph WaveNet model with random selection (30% dropout) of input FFT features outperforms other benchmarks with an AUROC of 88.57% to detect seizures. Random input selection effectively mitigates over-fitting to noise and promotes the identification of task-related frequencies through global exploration. This preprocessing strategy proves to be a simple yet effective method to improve model robustness, without prior knowledge and additional computational expense.

The weakly-supervised temporal action localization task aims to train a model that can accurately locate each action instance in the video using only video-level class labels. The existing methods take into account the information of different modalities (primarily RGB and Flow), and present numerous multi-modal complementary methods. RGB features are obtained by calculating appearance information, which are easy to be disrupted by the background. On the contrary, Flow features are obtained by calculating motion information, which are usually less disrupted by the background. Based on this phenomenon, we propose a Regional Similarity Consistency (RSC) constraint between these two modalities to suppress the disturbance of background in RGB features. Specifically, we calculate the regional similarity matrices of RGB and Flow features, and impose the consistency constraint through $$L_2$$ loss. To verify the effectiveness of our method, we integrate the proposed RSC constraint into three recent methods. The comprehensive experimental results show that the proposed RSC constraint can boost the performance of these methods, and achieve the state-of-the-art results on the widely-used THUMOS14 and ActivityNet1.2 datasets.

Weakly-supervised temporal action localization task aims to localize temporal boundaries of action instances by using only video-level labels. Existing methods primarily adopt Multi-Instance-Learning (MIL) scheme to handle this task. The effectiveness of MIL scheme depends heavily on the selection of top-k action snippets, which is unstable and requires manual tuning. To address these deficiencies, we propose an Adaptive Clustering and Refining Network (ACRNet). Specifically, we present an action-aware clustering strategy that is adaptable and requires no manual tuning to separate action and background snippets of diverse videos based on intra-class activation distribution. And a cluster refining step is included to eliminate false action snippets by considering inter-class activation distribution, which greatly improves robustness and localization accuracy. Extensive experiments on THUMOS14, ActivityNet 1.2&1.3 benchmarks show that our method achieves state-of-the-art performance.

Humans have the ability to appreciate and create music. However, why and how humans have this distinctive ability to perceive music remains unclear. Additionally, the investigation of the innate perceiving skill in humans is compounded by the fact that we have been actively and passively exposed to auditory stimuli or have systematically learnt music after birth. Therefore, to explore the innate musical perceiving ability, infants with preterm birth may be the most suitable population. In this study, the auditory brain networks were explored using dynamic functional connectivity-based reliable component analysis (RCA) in preterm infants during music listening. The brain activation was captured by portable functional near-infrared spectroscopy (fNIRS) to simulate a natural environment for preterm infants. The components with the maximum inter-subject correlation were extracted. The generated spatial filters identified the shared spatial structural features of functional brain connectivity across subjects during listening to the common music, exhibiting a functional synchronization between the right temporal region and the frontal and motor cortex, and synchronization between the bilateral temporal regions. The specific pattern is responsible for the functions involving music comprehension, emotion generation, language processing, memory, and sensory. The fluctuation of the extracted components and the phase variation demonstrates the interactions between the extracted brain networks to encode musical information. These results are critically important for our understanding of the underlying mechanisms of the innate perceiving skills at early ages of human during naturalistic music listening.

Surface electromyogram (sEMG)-based hand gesture recognition for prosthesis or armband is an important application of the human–machine interface (HMI). However, the measurement location of sensors greatly influences the hand gesture performance, especially with the interday or intersubject validation protocols. Therefore, we acquired two-day hand gesture data of 41 subjects with a 256 ( $16\times16$ ) channel high-density sEMG electrode array. With the acquired data, we initially compared the support vector machine (SVM) and other four state-of-art classifiers under three validation protocols, i.e., intraday, interday, and intersubject validation protocols. Then, we screened 14 feature optimization techniques, including five feature-projection methods and nine feature-ranking approaches. To present the accuracy tendency with varying measure locations, we systematically explored the ten-hand gesture performance using data of 16 prosthesis measurement locations (PMLs) and 15 armband measurement locations (AMLs). As a result, the SVM classifier was suitable for the intraday and interday validation protocols and the 2-D convolutional neural network was selected for the intersubject validation protocol. The mean accuracies of the hand gesture classification ranged from 95.68% to 99.12% (intraday validation), from 68.41% to 88.02% (interday validation), and from 63.39% to 86.33% (intersubject validation) for the prosthesis application. In addition, for the armband application, the mean accuracies ranged from 96.25% to 97.43% (intraday validation), from 67.44% to 75.83% (interday validation), and from 65.53% to 75.40% (intersubject validation). The accuracy is greatly correlated with the measurement location, which is highly associated with the neuromuscular structures of human bodies. In summary, our work can serve as a factor-screening tool for users customizing their systems according to their physical conditions and requirements.

Sleep research has evolved considerably since the first sleep electroencephalography (EEG) recordings in the 1930s and the discovery of well-distinguishable sleep stages in the 1950s. While electrophysiological recordings have been used to describe the sleeping brain in much detail, since the 1990s neuroimaging techniques are applied to uncover the brain organization and functional connectivity of human sleep with greater spatial resolution. The combination of EEG with different neuroimaging modalities such as Positron Emission Tomography (PET), structural MRI (sMRI) and functional Magnetic Resonance Imaging (fMRI) impose several challenges for sleep studies. For instance, difficulties maintaining and consolidating sleep in an unfamiliar and restricted environment, scanner-related distortions with physiological artifacts may contaminate polysomnography recordings, and the necessity to account for all physiological changes throughout the sleep cycles to better data interpretability. Here, we review the field of sleep neuroimaging in healthy non-sleep-deprived populations, from early findings to more recent developments, discuss the challenges of applying concurrent EEG and imaging techniques to sleep, and possible future directions the field will greatly benefit from.

Weakly-supervised temporal action localization seeks to localize temporal boundaries of actions while concurrently identifying their categories using only video-level category labels during training. Among the existing methods, the modal cooperation methods have achieved great success by providing pseudo supervision signals to RGB and Flow features. However, most of these methods ignore the cross-correlation between modal characteristics which can help them learn better features. By considering the cross-correlation, we propose a novel multi-head cross-modal attention mechanism to explicitly model the cross-correlation of modal features. The proposed method collaboratively enhances RGB and Flow features through a cross-correlation matrix. In this way, the enhanced features for each modality encode the inter-modal information, while preserving the exclusive and meaningful intra-modal characteristics. Experimental results on three recent methods demonstrate that the proposed Multi-head Cross-modal Attention (MCA) mechanism can significantly improve the performance of these methods, and even achieve state-of-the-art results on the THUMOS14 and ActivityNet1.2 datasets.

Music contains substantial contents that humans can perceive and thus has the capability to evoke positive emotions. Even though neonatal intensive care units (NICUs) can provide preterm infants a developmental environment, they still cannot fully simulate the environment in the womb. The reduced maternal care would increase stress levels in premature infants. Fortunately, music intervention has been proved that it can improve the NICU environment, such as stabilize the heart rate and the respiratory rate, reduce the incidence of apnea, and improve feeding. However, the effects of music therapy on the brain development of preterm infants need to be further investigated. In this paper, we evaluated the influence of short-term music therapy on the brain functions of preterm infants measured by functional near-infrared spectroscopy (fNIRS). We began by investigating how premature babies perceive structural information of music by calculating the correlations between music features and fNIRS signals. Then, the influences of short-term music therapy on brain functions were evaluated by comparing the resting-state functional connectivity before and after the short-term music therapy. The results show that distinct brain regions are responsible for processing corresponding musical features, indicating that preterm infants have the capability to process the complex musical content. However, the results of network analysis show that short-term music intervention is insufficient to cause the changes in cerebral functional connectivity. Therefore, long-term music therapy may be required to achieve the deserved effects on brain functional connectivity.

The fatigue-induced neuromuscular mechanism remains to be fully elucidated. So far, the macroscopic mechanism using global surface electromyogram (sEMG) has been widely investigated. However, the microscopic mechanism using high-level neural information based on motor unit (MU) spike train from the spinal cord lacks attention, especially for the conditions under dynamic contraction task. The synchronization of the MU spike train is generally assumed to be an excellent indicator to represent the activities of spinal nerves. Accordingly, this study employed synchronization of MU spike train decomposed from high-density sEMG (HD-sEMG) to investigate the fatigue condition in muscular contractions within the Biceps Brachii muscle under both isometric and dynamic contraction tasks, giving a complete picture of the microscopic fatigue mechanism. We compared the synchronization of MU in Delta (1-4 Hz), alpha (8-12 Hz), Beta (15-30 Hz), and Gamma (30-60 Hz) frequency bands during the fatigue condition induced by different contractions. Our results showed that MU synchronization increased significantly (p<0.05) in all frequency bands across the two contraction tasks. The results indicate that the microscopic fatigue mechanism of Biceps Brachii muscle does not vary due to different contraction tasks.

The ability to expertly control different fingers contributes to hand dexterity during object manipulation in daily life activities. The macroscopic spatial patterns of muscle activations during finger movements using global surface electromyography (sEMG) have been widely researched. However, the spatial activation patterns of microscopic motor units (MUs) under different finger movements have not been well investigated. The present work aims to quantify MU spatial activation patterns during movement of distinct fingers (index, middle, ring and little finger). Specifically, we focused on extensor muscles during extension contractions. Motor unit action potentials (MUAPs) during movement of each finger were obtained through decomposition of high-density sEMG (HD-sEMG). First, we quantified the spatial activation patterns of MUs for each finger based on 2-dimension (2-D) root-mean-square (RMS) maps of MUAP grids after spike-triggered averaging. We found that these activation patterns under different finger movements are distinct along the distal-proximal direction, but with partial overlap. Second, to further evaluate MU separability, we classified the spatial activation pattern of each individual MU under distinct finger movement and associated each MU with its corresponding finger with Regularized Uncorrelated Multilinear Discriminant Analysis (RUMLDA). A high accuracy of MU-finger classification tested on 12 subjects with a mean of 88.98% was achieved. The quantification of MU spatial activation patterns could be beneficial to studies of neural mechanisms of the hand. To the best of our knowledge, this is the first work which manages to quantify MU behaviors under different finger movements.

As a nondrug complementary therapy and healthy leisure physiotherapy method, foot bath (FB) is gaining acceptance and popularity in many areas. The significance of this research is to study the close and complex connection between FB stimulation and the human brain using fNIRS neuroimaging techniques. Participants were placed under two different conditions (normal and foot bath) and instructed to perform Stroop task of color word matching. Research on the behavioral results of the subjects showed that FB can effectively regulate the efficiency of humans in the process of performing tasks in a natural state. The fNIRS findings showed that the PFC in the FB condition was weakly activated compared to the normal condition. FB can realize the natural and healthy regulation of human brain cognitive function, which will have an impact on many production activities in human daily life.

Light modulates human brain function through its effect on circadian rhythms, which are related to several human behavioral and physiological processes. Functional near-infrared spectroscopy (fNIRS) is a noninvasive optical neuroimaging technique used for recording brain activation during task performance. This study aimed to investigate the effects of light on cognitive function, particularly in the prefrontal cortex using fNIRS. The effect of light on cognitive modulation was analyzed using the Stroop task, which was performed on 30 participants under three different light conditions (color temperature 4500 K, 2500 K, and none). The behavioral results indicated that light conditions can easily and effectively modulate the performance of tasks based on the feedback, including the response time and accuracy. fNIRS showed hemodynamic changes in the bilateral dorsolateral prefrontal cortices, and the activated brain regions varied under different light conditions. Moreover, light may be regarded as a safe, effective, inexpensive, and accessible tool for modulating human cognitive function.

Investigating cerebral hemodynamic changes during regular sleep cycles and sleep disorders is fundamental to understanding the nature of physiological and pathological mechanisms in the regulation of cerebral oxygenation during sleep. Although sleep neuroimaging methods have been studied and have been well-reviewed, they have limitations in terms of technique and experimental design. Neurologists are convinced that Near-infrared spectroscopy (NIRS) provides essential information and can be used to assist the assessment of cerebral hemodynamics, and numerous studies regarding sleep have been carried out based on NIRS. Thus, a brief historical overview of the sleep studies using NIRS will be helpful for the biomedical students, academicians, and engineers to better understand NIRS from various perspectives. In this study, the existing literature on sleep studies is reviewed, and an overview of the NIRS applications is synthesized and provided. The paper first reviews the application scenarios, as well as the patterns of fluctuation of NIRS, which includes the investigation in regular sleep and sleep-disordered breathing. Various factors such as different sleep stages, populations, and degrees of severity were considered. Furthermore, the experimental design and signal processing, as well as the regulation mechanisms involved in regular and pathological sleep, are investigated and discussed. The strengths and weaknesses of the existing NIRS applications are addressed and presented, which can direct further NIRS analysis and utilization.

Acting as a brain stimulant, coffee resulted in heightening alertness, keeping arousal, improving executive speed, maintaining vigilance, and promoting memory, which are associated with attention, mood, and cognitive function. Functional near-infrared spectroscopy (fNIRS) is a noninvasive optical method to monitor brain activity by measuring the absorption of the near-infrared light through the intact skull. This study is aimed at acquiring brain activation during executing task performance. The aim is to explore the effect of coffee on cognitive function by the fNIRS neuroimaging method, particularly on the prefrontal cortex regions. The behavioral experimental results on 31 healthy subjects with a Stroop task indicate that coffee can easily and effectively modulate the execute task performance by feedback information of the response time and accuracy rate. The findings of fNIRS showed that apparent hemodynamic changes were detected in the bilateral VLPFC regions and the brain activation regions varied with different coffee conditions.

10 BACKGROUND: Stroke is a leading cause of mortality and disability, people’s daily living habits can affect the risk of stroke. 11 OBJECTIVE: To investigate the effects of main daily living habits (smoking, drinking, diet, vegetable and fruits consumption, 12 and exercise) on stroke risk in patients and provide the scientific basis for the assessment of the risk factors, a novel risk analysis 13 model of the stroke. 14 METHODS: A data mining method using decision trees which adopted the optimized C4.5 algorithm is presented, which are 15 derived from a series of 23682 patients with 21 risk factors. 16 RESULTS: The overall accuracy and kappa coefficient for stroke risk classification has reached 84.88% and 0.7763, respec17 tively. Through the generated knowledge rules, it demonstrates that the behavioral habits in daily life have an indirect effect on 18 the risk of stroke. While, it has an obvious effect on stroke when hypertension, diabetes mellitus, hypercholesterolemia, and 19 BMI risk factors exist. In addition, it was observed that the aforementioned five daily living habits have a decreased impact on 20 the stroke. 21 CONCLUSIONS: It is anticipated that the proposed system could help in reducing the risk, mortality, and disability of stroke, 22 and provide clinical decision support for the treatment of stroke. 23

Preterm infants, as a special population, exposed to the stimulus of medical procedures, need efficient integrated care. Music, a non-invasive positive stimulus, has the ability to improve infants well-being. In this study, a randomized controlled trial was conducted to evaluate the effect of Mozart music on cerebral hemodynamics detected by functional near-infrared spectroscopy (fNIRS). The statistical analysis results showed that the sample entropy of the experimental group was significantly lower than that of the control group in resting state after music intervention. It suggested that Mozart music could achieve the effect of calmness and relaxed state correlated with lower brain activities. This study will provide evidence to support the wide application of music intervention in clinical practice.

BACKGROUND: Stroke is a leading cause of mortality and disability, which can be affected by people’s daily living habits. OBJECTIVE: To investigate the effects of main daily living habits (smoking, drinking, diet, vegetable and fruits consumption, and exercise) on stroke risk in patients and provide the scientific basis for the assessment of the risk factors, a novel risk analysis model of the stroke is proposed. METHODS: A data mining method using decision trees which adopted the optimized C4.5 algorithm is presented. It is able to deal with the unbalanced data problem of the classification. Meanwhile, the proposed method has been verified on a clinical dataset of 23,682 patients with 21 risk factors. RESULTS: The overall accuracy and kappa coefficient for stroke risk classification has reached 84.88% and 0.7763, respectively. Through the generated knowledge rules, it demonstrates that the behavioral habits in daily life have an indirect effect on the risk of stroke. While, it has an obvious effect on stroke when hypertension, diabetes mellitus, hypercholesterolemia, and BMI risk factors exist. In addition, it was observed that the aforementioned five daily living habits have a decreased impact on the stroke. CONCLUSIONS: It is anticipated that the proposed system could help in reducing the risk, mortality, and disability of stroke, and provide clinical decision support for the treatment of stroke.