Spiking neural networks (SNNs) offer a promising path toward energy-efficient speech command recognition (SCR) by leveraging their event-driven processing paradigm. However, existing SNN-based SCR methods often struggle to capture rich temporal dependencies and contextual information from speech due to limited temporal modeling and binary spike-based representations. To address these challenges, we first introduce the multi-view spiking temporal-aware self-attention (MSTASA) module, which combines effective spiking temporal-aware attention with a multi-view learning framework to model complementary temporal dependencies in speech commands. Building on MSTASA, we further propose SpikCommander, a fully spike-driven transformer architecture that integrates MSTASA with a spiking contextual refinement channel MLP (SCR-MLP) to jointly enhance temporal context modeling and channel-wise feature integration. We evaluate our method on three benchmark datasets: the Spiking Heidelberg Dataset (SHD), the Spiking Speech Commands (SSC), and the Google Speech Commands V2 (GSC). Extensive experiments demonstrate that SpikCommander consistently outperforms state-of-the-art (SOTA) SNN approaches with fewer parameters under comparable time steps, highlighting its effectiveness and efficiency for robust speech command recognition.

Emotion Recognition in Conversation (ERC) is valuable for diagnosing health conditions such as autism and depression (Maryenko, 2024), and for understanding the emotions of individuals who struggle to express their feelings. Current ERC methods primarily rely on semantic, audio and video data but face significant challenges in integrating physiological signals such as Electroencephalography (EEG), which has low signal-to-noise ratios, inter-subject variability, and temporal alignment issues. This research proposes Hypergraph Multi-Modal Learning (Hyper-MML), a novel framework for identifying emotions in conversation. Hyper-MML effectively integrates EEG with audio and video information to capture complex emotional dynamics. Firstly, we introduce an Adaptive Brain Encoder with Mutual-cross Attention (ABEMA) module for processing EEG signals. This module captures emotion-relevant features across different frequency bands and adapts to subject-specific variations through hierarchical mutual-cross attention mechanisms. Secondly, we propose an Adaptive Hypergraph Fusion Module (AHFM) to actively model the higher-order relationships among multi-modal signals in ERC. Experimental results on the EAV and AFFEC datasets demonstrate that our Hyper-MML model significantly outperforms current state-of-the-art methods. The proposed Hyper-MML can serve as an effective communication tool for healthcare professionals, enabling better engagement with patients who have difficulty expressing their emotions. The official implementation codes are available at https://github.com/NZWANG/Hyper-MML.

Effective regulation of social decision-making is crucial for achieving equitable outcomes in human interactions. This study explores the impact of endogenous regulation on social decision-making and associated neural changes through a neurofeedback (NF) training framework. Given the relationship between social decision making, emotions, and frontal alpha asymmetry (FAA), this NF training enables individuals to self-regulate their FAA, thereby influencing their decision-making behavior. Eighty-one participants were randomly divided into the up-FAA group aiming at up-regulating FAA, the down-FAA group aiming at down-regulating FAA, and the sham-NF group. First, our results validated the specific NF training effect on selfregulating FAA. Notably, not all participants in the up-FAA and down-FAA groups successfully learned to regulate their FAA, leading to further subdivision into up-learner, down-learner, up-nonlearner, and down-nonlearner categories based on learning efficacy. Participants who effectively learned to reduce their FAA (down-learners) showed significant changes in decision behavior under moderately unfair conditions, characterized by increased rejection rates during the ultimatum game (UG) task. They also exhibited larger N200 amplitudes while balancing the decisionmaking period. In contrast, up learners demonstrated minimal behavioral changes despite increases in FAA. We conclude that decreases in FAA have a more pronounced impact on social decision-making than increases during NF training. This study highlights the effects of FAA self-regulation on fairness-related decision-making, revealing the neurobiological factors that shape decisions influenced by fairness perceptions. These findings offer valuable insights for enhancing social cooperation and justice.

Brain-computer interface (BCI) based on steady- state visual evoked potentials (SSVEP) is a popular paradigm for its simplicity and high information transfer rate (ITR). Accurate and fast SSVEP decoding is crucial for reliable BCI performance. However, conventional decoding methods demand longer time windows, and deep learning models typically require subject-specific fine-tuning, leaving challenges in achieving optimal performance in cross-subject settings. This paper proposed a biofocal masking attention-based method (SSVEP-BiMA) that synergistically leverages the native and symmetric-antisymmetric components for decoding SSVEP. By utilizing multiple signal representations, the network is able to integrate features from a wider range of sample perspectives, leading to more generalized and comprehensive feature learning, which enhances both prediction accuracy and robustness. We performed experiments on two public datasets, and the results demonstrate that our proposed method surpasses baseline approaches in both accuracy and ITR. We believe that this work will contribute to the development of more efficient SSVEP-based BCI systems.

The development of EEG decoding algorithms confronts challenges such as data sparsity, subject variability, and the need for precise annotations, all of which are vital for advancing brain-computer interfaces and enhancing the diagnosis of diseases. To address these issues, we propose a novel two-stage approach named Self-Supervised State Reconstruction-Primed Riemannian Dynamics (EEG-ReMinD), which mitigates reliance on supervised learning and integrates inherent geometric features. This approach efficiently handles EEG data corruptions and reduces the dependency on labels. EEG-ReMinD utilizes self-supervised and geometric learning techniques, along with an attention mechanism, to analyze the temporal dynamics of EEG features within the framework of Riemannian geometry, referred to as Riemannian dynamics. Comparative analyses on both intact and corrupted datasets from two different neurodegenerative disorders underscore the enhanced performance of EEG-ReMinD.

Auditory attention detection (AAD) aims to decode listeners' focus in complex auditory environments from electroencephalography (EEG) recordings, which is crucial for developing neuro-steered hearing devices. Despite recent advancements, EEG-based AAD remains hindered by the absence of synergistic frameworks that can fully leverage complementary EEG features under energy-efficiency constraints. We propose ***S$^2$M-Former***, a novel ***s***piking ***s***ymmetric ***m***ixing framework to address this limitation through two key innovations: i) Presenting a spike-driven symmetric architecture composed of parallel spatial and frequency branches with mirrored modular design, leveraging biologically plausible token-channel mixers to enhance complementary learning across branches; ii) Introducing lightweight 1D token sequences to replace conventional 3D operations, reducing parameters by 14.7$\times$. The brain-inspired spiking architecture further reduces power consumption, achieving a 5.8$\times$ energy reduction compared to recent ANN methods, while also surpassing existing SNN baselines in terms of parameter efficiency and performance. Comprehensive experiments on three AAD benchmarks (KUL, DTU and AV-GC-AAD) across three settings (within-trial, cross-trial and cross-subject) demonstrate that S$^2$M-Former achieves comparable state-of-the-art (SOTA) decoding accuracy, making it a promising low-power, high-performance solution for AAD tasks. Code is available at https://github.com/JackieWang9811/S2M-Former.

Prediction of daily mental wellbeing holds profound implications for individual healthcare and societal stability. Previous studies have shown the potential of using individual's multimodal behavioral data collected through mobile devices to predict his/her daily mental wellbeing metrics, such as stress, mood, and anxiety. However, effectively capturing long-range dependencies in behavioral time series data while accurately representing the statistical distribution patterns of various behaviors over a certain period is a significant challenge. In this paper, we propose a daily mental wellbeing prediction model based on a Topic-Guided Self-Attention Network (TGSAN). This model utilizes self-attention mechanism to capture long-range dependencies from the behavioral data collected by mobile devices. We utilize a multi-granularity time encoding method to inject time information of different granularities (i.e., day and hour, or week and day) into the behavioral data, thereby enhancing the sensibility of the self-attention network to capture every individual's habitual cyclicality rhythm. Then, we introduce a neural topic model to analyze the statistical distribution characteristics of various behaviors in the monitoring period as behavioral distribution patterns for different individuals, and further propose a topic attention network to enhance the model's classification performance by guiding the weights of long-range dependencies features from the self-attention network with the derived topic information. Compared to state-of-the-art methods, the proposed TGSAN achieved superior performance on datasets that measure different mental health indicators (stress, mood, and anxiety), with F1 scores outperforming by 4.5% and 2.3% on the Crosscheck and StudentLife datasets, respectively, and accuracy outperforming by 3.3% on the GLOBEM dataset. Our study demonstrates the effectiveness and interpretability of combining self-attention mechanisms with neural topic model, for a better understanding of the relationship between different individuals’ behaviors and their mental wellbeing.

Electroencephalography (EEG) is an objective tool for emotion recognition with promising applications. However, the scarcity of labeled data remains a major challenge in this field, limiting the widespread use of EEG-based emotion recognition. In this paper, a semi-supervisedDual-streamSelf-attentiveAdversarialGraphContrastive learning framework (termed asDS-AGC) is proposed to tackle the challenge of limited labeled data in cross-subject EEG-based emotion recognition. The DS-AGC framework includes two parallel streams for extracting non-structural and structural EEG features. The non-structural stream incorporates a semi-supervised multi-domain adaptation method to alleviate distribution discrepancy among labeled source domain, unlabeled source domain, and unknown target domain. The structural stream develops a graph contrastive learning method to extract effective graph-based feature representation from multiple EEG channels in a semi-supervised manner. Further, a self-attentive fusion module is developed for feature fusion, sample selection, and emotion recognition, which highlights EEG features more relevant to emotions and data samples in the labeled source domain that are closer to the target domain. Extensive experiments are conducted on four benchmark databases (SEED, SEED-IV, SEED-V, and FACED) using a semi-supervised cross-subject leave-one-subject-out cross-validation evaluation protocol. The results show that the proposed model outperforms existing methods under different incomplete label conditions with an average improvement of 2.17%, which demonstrates its effectiveness in addressing the label scarcity problem in cross-subject EEG-based emotion recognition.

Electroencephalography (EEG) is an objective tool for emotion recognition and shows promising performance. However, the label scarcity problem is a main challenge in this field, which limits the wide application of EEG-based emotion recognition. In this article, we propose a novel semisupervised transfer learning framework (EEGMatch) to leverage both labeled and unlabeled EEG data. First, an EEG-Mixup-based data augmentation method is developed to generate more valid samples for model learning. Second, a semisupervised two-step pairwise learning method is proposed to bridge prototypewise and instancewise pairwise learning, where the prototypewise pairwise learning measures the global relationship between EEG data and the prototypical representation of each emotion class and the instancewise pairwise learning captures the local intrinsic relationship among EEG data. Third, a semisupervised multidomain adaptation is introduced to align the data representation among multiple domains (labeled source domain, unlabeled source domain, and target domain), where the distribution mismatch is alleviated. Extensive experiments are conducted on three benchmark databases (SEED, SEED-IV, and SEED-V) under a cross-subject leave-one-subject-out cross-validation evaluation protocol. The results show the proposed EEGMatch performs better than the state-of-the-art methods under different incomplete label conditions (with 5.89% improvement on SEED, 0.93% improvement on SEED-IV, and 0.28% improvement on SEED-V), which demonstrates the effectiveness of the proposed EEGMatch in dealing with the label scarcity problem in emotion recognition using EEG signals. The source code is available at https://github.com/KAZABANA/EEGMatch.

Reconstructing natural language from noninvasive electroencephalography (EEG) holds great promise as a language decoding technology for brain-computer interfaces (BCIs).How-

Identifying mild cognitive impairment (MCI) is vital for Alzheimer’s disease prevention. As neurodegenerative diseases progress, synchronous activity in electroencephalography (EEG) - indicating functional connectivity - changes due to neural system deterioration. Thus, developing geometric learning to decode the functional brain structure is essential. Techniques such as graph neural networks and Riemannian manifolds show potential in analyzing non-Euclidean data. However, existing approaches neglect to combine synchronous activity with temporal dependence and still remain insufficient for MCI detection. This paper proposes the Brain Network sequence-driven Manifold-based Transformer (BNMTrans) to identify MCI patterns from EEG data. BNMTrans leverages its strengths by extracting features from sequential brain networks through the self-attention mechanism, guided by the geometric correlations within the Riemannian manifold. By integrating long-term temporal dynamics and structural relationships within manifold space based on functional connectivity, this approach outperforms others in EEG feature comparisons and state-of-the-art evaluations based on clinical data from 89 subjects (46 MCI, 43 healthy controls) at a local hospital. Our work has significance for both MCI clinical management and technical progression in the EEG field.

Automatic depression detection based on audio and text representations from participants’ interviews has attracted widespread attention. However, most of previous researches only used one type of feature of one single modality for depression detection, so that the rich information of audio and text from interviews has not been fully utilized. Moreover, an effective multi-modal fusion approach to leverage the independence among audio and text representations is still lacking. To address these problems, we propose a multi-modal fusion depression detection model based on the interaction of multilevel audio features and text sentence embedding. Specifically, we first extract Low-Level Descriptors (LLDs), mel-spectrogram features, and wav2vec features from the audio. Then we design a Multi-level Audio Features Interaction Module (MAFIM) to fuse these three levels of features for a comprehensive audio representation. For interview text, we use pre-trained BERT to extract sentence-level embedding. Further, to effectively fuse audio and text representations, we design a Channel Attention-based Multi-modal Fusion Module (CAMFM) by taking into account the independence and correlation between two different modalities. Our proposed model shows better performance on two datasets, DAIC-WOZ and EATD-Corpus, than existing methods, so it has a high potential to be applied for interview-based depression detection in practice.

Emotion recognition from electroencephalography (EEG) has attracted widespread interest, but few studies have considered estimating the highly dynamic trajectories of emotion in a relatively long period, such as video watching. To address this problem, we first recruit participants to assign continuous-time and continuous-level emotion labels to videos from the SEED corpus. Then, we propose a hybrid model, namely Emotion Time-Varying Regression (EmoTVR), to estimate continuous-time and continuous-level emotion using EEG spatial-temporal representations. EmoTVR combines atrous convolutional networks for spatial feature extraction and a temporal self-attentive regressor using attention-based long short-term memory for temporal feature extraction and continuous estimation. Moreover, EmoTVR adopts the Domain Adversarial Neural Network to address the problem of individual difference. Experimental results through both within-subject and cross-subject cross-validations demonstrate the superiority of EmoTVR in recognizing dynamic and continuous emotion over traditional methods. The proposed EmoTVR method caters for the needs of dynamic and continuous emotion recognition in naturalistic conditions, so it is highly potential for practical applications of emotion recognition.

Cross-center data heterogeneity and annotation unreliability significantly challenge the intelligent diagnosis of diseases using brain signals. A notable example is the EEG-based diagnosis of neurodegenerative diseases, which features subtler abnormal neural dynamics typically observed in small-group settings. To advance this area, in this work, we introduce a transferable framework employing Manifold Attention and Confidence Stratification (MACS) to diagnose neurodegenerative disorders based on EEG signals sourced from four centers with unreliable annotations. The MACS framework's effectiveness stems from these features: 1) The Augmentor generates various EEG-represented brain variants to enrich the data space; 2) The Switcher enhances the feature space for trusted samples and reduces overfitting on incorrectly labeled samples; 3) The Encoder uses the Riemannian manifold and Euclidean metrics to capture spatiotemporal variations and dynamic synchronization in EEG; 4) The Projector, equipped with dual heads, monitors consistency across multiple brain variants and ensures diagnostic accuracy; 5) The Stratifier adaptively stratifies learned samples by confidence levels throughout the training process; 6) Forward and backpropagation in MACS are constrained by confidence stratification to stabilize the learning system amid unreliable annotations. Our subject-independent experiments, conducted on both neurocognitive and movement disorders using cross-center corpora, have demonstrated superior performance compared to existing related algorithms. This work not only improves EEG-based diagnostics for cross-center and small-setting brain diseases but also offers insights into extending MACS techniques to other data analyses, tackling data heterogeneity and annotation unreliability in multimedia and multimodal content understanding. We have released our code here: https://github.com/ICI-BCI/EEG-MACS.

Pain is highly subjective, so it is always desirable to develop objective pain assessment methods. Brain imaging techniques, such as functional magnetic resonance imaging (fMRI), have the potential to provide a physiological and quantitative pain assessment tool. However, the ultra-high-dimensional fMRI data and the nonlinear relationship between fMRI and pain greatly degrade the efficiency of fMRI-based pain decoding models. In this paper, a novel pain decoding model is proposed based on the whale optimization algorithm (WOA), reinforcement learning (RL), and self-organizing fuzzy logic (SOFL), namely RLWOA-SOFL. The new non-linear WOA method incorporates RL and repository experiences (RE), which is based on a back-propagation neural network (BPNN) to map a set of agents states to appropriate actions, to extract and select features that are highly predictive of pain. More specifically, the proposed RLWOA is self-learning and self-optimizing so it can deal with the high-dimensional and complex fMRI data. On the other hand, to establish a fMRI-based pain decoding model, a novel SOFL method is proposed as a new type of deep fuzzy rule that can learn continuously from new data and identify prototypes to construct fuzzy rules. The proposed RLWOA-SOFL model is applied to real-world pain-evoked fMRI data, and the results show that the new model can decode pain intensity more accurately and can identify pain-related fMRI patterns more reliably. Therefore, the proposed RLWOA-SOFL model has great potential to evaluate the intensity of pain perception in clinical uses.

Recognizing human emotions from complex, multivariate, and non-stationary electroencephalography (EEG) time series is essential in affective brain-computer interface. However, because continuous labeling of ever-changing emotional states is not feasible in practice, existing methods can only assign a fixed label to all EEG timepoints in a continuous emotion-evoking trial, which overlooks the highly dynamic emotional states and highly non-stationary EEG signals. To solve the problems of high reliance on fixed labels and ignorance of time-changing information, in this paper we propose a time-aware sampling network (TAS-Net) using deep reinforcement learning (DRL) for unsupervised emotion recognition, which is able to detect key emotion fragments and disregard irrelevant and misleading parts. Specifically, we formulate the process of mining key emotion fragments from EEG time series as a Markov decision process and train a time-aware agent through DRL without label information. First, the time-aware agent takes deep features from a feature extractor as input and generates sample-wise importance scores reflecting the emotion-related information each sample contains. Then, based on the obtained sample-wise importance scores, our method preserves top-Xcontinuous EEG fragments with relevant emotion and discards the rest. Finally, we treat these continuous fragments as key emotion fragments and feed them into a hypergraph decoding model for unsupervised clustering. Extensive experiments are conducted on three public datasets (SEED, DEAP, and MAHNOB-HCI) for emotion recognition using leave-one-subject-out cross-validation, and the results demonstrate the superiority of the proposed method against previous unsupervised emotion recognition methods. The proposed TAS-Net has great potential in achieving a more practical and accurate affective brain-computer interface in a dynamic and label-free circumstance. The source code is made available athttps://github.com/infinite-tao/TAS-Net.

Affective brain-computer interface based on electroencephalography (EEG) is an important branch in the field of affective computing. However, the individual differences in EEG emotional data and the noisy labeling problem in the subjective feedback seriously limit the effectiveness and generalizability of existing models. To tackle these two critical issues, we propose a novel transfer learning framework with Prototypical Representation based Pairwise Learning (PR-PL). The discriminative and generalized EEG features are learned for emotion revealing across individuals and the emotion recognition task is formulated as pairwise learning for improving the model tolerance to the noisy labels. More specifically, a prototypical learning is developed to encode the inherent emotion-related semantic structure of EEG data and align the individuals' EEG features to a shared common feature space under consideration of the feature separability of both source and target domains. Based on the aligned feature representations, pairwise learning with an adaptive pseudo labeling method is introduced to encode the proximity relationships among samples and alleviate the label noises effect on modeling. Extensive results on two benchmark databases (SEED and SEED-IV) under four different cross-validation evaluation protocols validate the model reliability and stability across subjects and sessions. Compared to the literature, the average enhancement of emotion recognition across four different evaluation protocols is 2.04% (SEED) and 2.58% (SEED-IV). The source code is available athttps://github.com/KAZABANA/PR-PL.

The diagnosis of cognitive impairment (CI), here referred to as mild cognitive impairment (MCI) and probable Alzheimer’s disease (AD), is complicated in practice. Early AD diagnosis using electroencephalography (EEG) has attracted attention due to EEG’s advantages in data accessibility. Because of limited, sparse, and ambiguous labels, which are commonly encountered in the EEG-based diagnosis of CI, it is desirable to develop a learning framework to effectively capture CI-related representations beyond fully supervised learning. Therefore, this work explored the possibility of weakly-supervised learning in identifying MCI, AD, and normal aging patterns based on incompletely reliable labels. To address the problem, we proposed a framework containing a dual-contrastive learning structure and a multi-level temporal-spectral EEG encoder, which transformed EEG signals into embeddings and automatically updated the ambiguous labels through intra-subject and cross-subject contrastive learning. We verified the method’s performance based on 54 subjects (18 in each group). Our findings provide new insights into the accurate inference of refractory CI diseases based on non-ideal data sources.

Machine learning has been increasingly used in decoding brain states from functional magnetic resonance imaging (fMRI). One important application is to classify the levels of pain perception from patients’ fMRI for clinical pain assessment. However, the huge number of fMRI features and the complex relationships between fMRI and pain levels affect the performance of pain classification models heavily. In this article, we introduce a new fuzzy-rule-based hybrid optimization approach for dimension reduction and multiclassification problems using chaotic map, crow search optimization (CSO), and self-organizing fuzzy logic prototype (SOFLP). The approach is named as CCSO–SOFLP. In the proposed approach, chaotic map-based CSO is employed to find the optimal features from ultra-high-dimensional fMRI, and the fuzzy-rule-based SOFLP is employed for multiclassification of pain levels. In this sense, CSO is provided to avoid being stuck in local minima and to increase the computational performance. On the other hand, multilayer SOFLP classifier can continuously learn from new data and identify prototypes from the observed data and use them to build fuzzy rules, to define a suitable local area for each prototype, and to avoid overlapping. The proposed approach is applied on a pain-evoked fMRI data set to classify the levels of pain. Results indicate that the proposed approach can decode levels of pain and identify predictive fMRI patterns with higher accuracy and convergence speed and shorter execution time. Therefore, the new type of fuzzy-rule-based system with chaotic swarm intelligence holds great potential to predict pain perception in clinical uses.

Training gene expression data with supervised learning approaches can provide an alarm sign for early treatment of lung cancer to decrease death rates. However, the samples of gene features involve lots of noises in a realistic environment. In this study, we present a random forest with self-paced learning bootstrap for improvement of lung cancer classification and prognosis based on gene expression data. To be specific, we propose an ensemble learning with random forest approach to improving the model classification performance by selecting multi-classifiers. Then, we investigate the sampling strategy by gradually embedding from high- to low-quality samples by self-paced learning. The experimental results based on five public lung cancer datasets show that our proposed method could select significant genes exactly, which improves classification performance compared to that of existing approaches. We believe that our proposed method has the potential to assist doctors in gene selections and lung cancer prognosis.

This paper proposes a new algorithm for automatic estimation of muscle fiber orientation (MFO) in musculoskeletal ultrasound images, which is commonly used for both diagnosis and rehabilitation assessment of patients. The algorithm is based on a novel adaptive fading Bayesian Kalman filter (AF-BKF) and an automatic region of interest (ROI) extraction method. The ROI is first enhanced by the Gabor filter (GF) and extracted automatically using the revoting constrained Radon transform (RCRT) approach. The dominant MFO in the ROI is then detected by the RT and tracked by the proposed AF-BKF, which employs simplified Gaussian mixtures to approximate the non-Gaussian state densities and a new adaptive fading method to update the mixture parameters. An AF-BK smoother (AF-BKS) is also proposed by extending the AF-BKF using the concept of Rauch-Tung-Striebel smoother for further smoothing the fascicle orientations. The experimental results and comparisons show that: 1) the maximum segmentation error of the proposed RCRT is below nine pixels, which is sufficiently small for MFO tracking; 2) the accuracy of MFO gauged by RT in the ROI enhanced by the GF is comparable to that of using multiscale vessel enhancement filter-based method and better than those of local RT and revoting Hough transform approaches; and 3) the proposed AF-BKS algorithm outperforms the other tested approaches and achieves a performance close to those obtained by experienced operators (the overall covariance obtained by the AF-BKS is 3.19, which is rather close to that of the operators, 2.86). It, thus, serves as a valuable tool for automatic estimation of fascicle orientations and possibly for other applications in musculoskeletal ultrasound images.

Studies of time-varying or dynamic brain connectivity (BC) using functional magnetic resonance imaging (fMRI) are crucial to understand the relationship between different brain regions. This paper presents a novel method for estimating dynamic BC using a time-varying multivariate autoregressive (AR) model with spatial sparsity and temporal continuity constraints. The problem is formulated as a maximum a posterior probability (MAP) estimation problem and solved as a least square problem with Li-regularization for imposing the constraints. The Limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS) method is employed to estimate the model parameters for making inference of dynamic BC. The proposed method was evaluated using synthetic data and visual checkerboard task experiment fMRI data. The results show that the method can effectively capture transient information transfer among visual-related brain regions whereas controlled areas not related to the process remain inactive. These verify the effectiveness and reduced variance of the proposed method for investigating dynamic task-related BC from fMRI data.

In this study, an automatic muscle fiber orientation tracking approach based on Bayesian Kalman Filter (BKF) is proposed. The BKF employs a Gaussian mixture (GM) representation of the state and noise densities and a novel direct density simplifying algorithm for avoiding the exponential complexity growth of conventional Kalman filters (KFs) using GM. In this paper, the ultrasound image is firstly enhanced by a bank of Gabor Filters (GFs) based on the GM of the state density in BKF. Then, a bank of localized radon transforms (LRTs) are used to extract muscle fiber orientations and the dominant orientation is obtained by minimizing an energy function. Finally, the dominant orientation is fed back to the BKF as an observation. The performance of the proposed approach is compared with existing methods on five subjects over 1000+ clinical ultrasound images. Experimental results show that the proposed method can achieve accurate and robust measurements of fascicle orientation and outperforms all the existing methods.

This paper is aimed to predict pain perception from laser-evoked EEG oscillatory activities in the time-frequency domain with multivariate pattern analysis (MVPA). We first identify pre-/post-stimulus EEG oscillatory activities that are correlated with the intensity of laser-evoked pain perception using a multivariate linear regression (MVLR) model, which is solved by partial least-squares regression (PLSR). Further, we used the MVLR model to predict the intensity of pain perception from identified pain-correlated time-frequency EEG data for each subject. Our results showed that the proposed MVLR prediction model provided a qualitative prediction of pain (classification of low pain and high pain) with an accuracy of 78.53 ± 1.16% and a quantitative prediction of pain (on a continuous scale from 0 to 10) with a mean absolute error (MAE) of 1.45 ± 0.05, both of which are significantly better than the results of the conventional pain prediction based on single-trial detection of laser-evoked potentials. Besides, for the first time it was found that the pre-stimulus EEG oscillation could significantly contribute to the prediction, which extended our notion of the determinants of pain perception.

The study is aimed to investigate in vivo behaviors of the rectus femoris muscle during isometric contraction by integrating simultaneously recorded electromyography (EMG), mechanomyography (MMG), and ultrasonography (US). We developed an experimental platform for simultaneous acquisition of EMG, MMG, US, as well as the torque, during isometric muscle contraction. Features from multimodal signals and images were then automatically extracted and calibrated to present time-varying characteristics of muscle behaviors. We further applied local polynomial regression (LPR) to reveal nonlinear and transient relationships between multimodal muscle features and torque. The results suggested that the proposed multimodal signal acquisition and integration are capable of providing novel and complete information about in vivo muscle contraction. The proposed experimental platform is a potentially useful tool for muscle assessment in various clinical and practical applications.

This paper introduces a new method for adaptively estimating the time-varying autocorrelation (TV-AC) of nonstationary signals and studies its application to time-frequency analysis. The proposed method employs local estimation with a sliding window having a certain bandwidth to estimate the TV-AC locally. The window bandwidths are selected adaptively by a local plug-in rule to address the bias and variance tradeoff problem. Further, based on the proposed adaptive TV-AC estimation, a new time-frequency analysis method called adaptive windowed minimum variance spectral estimation (AWMVSE) is developed. Simulation results show that the proposed adaptive TV-AC estimation method and AWMVSE method have improved performances over conventional estimators with a fixed window.

This paper proposes a new regularized transform domain normalized LMS (R-TDNLMS) algorithm and studies its mean and mean square convergence performances. The proposed algorithm extends the conventional TDNLMS algorithm by imposing a regularization term on the filter coefficients to reduce the variance of estimators due to the lacking of excitation in a certain frequency band or in the presence of modeling errors. Difference equations describing the mean and mean square convergence behaviors of this algorithm are derived so as to characterize its convergence condition and steady-state excess mean square error (MSE). It shows that regularization can help to reduce the MSE by trading slight bias for variance. Based on this analysis, a new formula to select the regularization parameter for white Gaussian inputs is proposed, which leads to a new variable regularized TDNLMS (VR-TDNLMS) algorithm. Computer simulations are conducted to examine the improved convergence performance, steady-state MSE and robustness to power-varying inputs of the proposed algorithm and verify the effectiveness of the theoretical analysis. Furthermore, the application of the proposed VR-TDNLMS algorithm to the design and implementation of acoustic system identification and active noise control (ANC) systems show that they considerably outperforms traditional TDNLMS algorithms at low excitation or in the presence of modeling errors. Moreover, the theoretical analysis provides simple design formulas for achieving a given excess MSE (EMSE) and step-size bound for stable operation.

This paper proposes a new variable regularized QR decompPosition (QRD)-based recursive least M-estimate (VR-QRRLM) adaptive filter and studies its convergence performance and acoustic applications. Firstly, variableL2regularization is introduced to an efficient QRD-based implementation of the conventional RLM algorithm to reduce its variance and improve the numerical stability. Difference equations describing the convergence behavior of this algorithm in Gaussian inputs and additive contaminated Gaussian noises are derived, from which new expressions for the steady-state excess mean square error (EMSE) are obtained. They suggest that regularization can help to reduce the variance, especially when the input covariance matrix is ill-conditioned due to lacking of excitation, with slightly increased bias. Moreover, the advantage of the M-estimation algorithm over its least squares counterpart is analytically quantified. For white Gaussian inputs, a new formula for selecting the regularization parameter is derived from the MSE analysis, which leads to the proposed VR-QRRLM algorithm. Its application to acoustic path identification and active noise control (ANC) problems is then studied where a new filtered-x (FX) VR-QRRLM ANC algorithm is derived. Moreover, the performance of this new ANC algorithm under impulsive noises and regularization can be characterized by the proposed theoretical analysis. Simulation results show that the VR-QRRLM-based algorithms considerably outperform the traditional algorithms when the input signal level is low or in the presence of impulsive noises and the theoretical predictions are in good agreement with simulation results.

This paper proposes a new state-regularized (SR) and QR decomposition based recursive least squares (QRRLS) algorithm with variable forgetting factor (VFF) for recursive coefficient estimation of time-varying autoregressive (AR) models. It employs the estimated coefficients as prior information to minimize the exponentially weighted observation error, which leads to reduced variance and bias over traditional regularized RLS algorithm. It also increases the tracking speed by introducing a new measure of convergence status to control the FF. Simulations using synthetic and real speech signals show that the proposed method has improved tracking performance and reduced estimation error variance than conventional TVAR modeling methods during rapid changing of AR coefficients.

The transform domain normalized least mean squares (TDNLMS) algorithm is an efficient adaptive algorithm, which offers fast convergence speed with a reasonably low arithmetic complexity. However, its convergence speed is usually limited by the fixed step-size so as to achieve a low desired misadjustment. In this paper a new switch-mode noise-constrained TDNLMS (SNC-TDNLMS) algorithm is proposed. It employs a maximum step-size mode in initial convergence and a noise-constrained mode afterwards to improve the convergence speed and steady state performance. The mean and mean square convergence behaviors of the proposed algorithm are studied to characterize its convergence condition and steady-state excess mean square error (EMSE). Based on the theoretical results, an automatic threshold selection scheme for mode switching is developed. Computer simulations are conducted to show the effectiveness of the proposed algorithm and verify the theoretical results.

This paper develops a subspace-based method for direction-of-arrival (DOA) estimation of uniform linear array (ULA) in the presence of mutual coupling. As the mutual coupling coefficient between two sensor elements is inversely related to their separation and is negligible when they are separated by a few wavelengths, the mutual coupling matrix (MCM) of a ULA can be well approximated as a banded symmetric Toeplitz matrix, which greatly reduces the number of unknown parameters to be estimated. Using the subspace principle, we propose a new method for joint estimation of the DOAs of incoming signals and banded symmetric Toeplitz MCM by reconstructing the steering vector to a specific matrix form. The proposed method achieves a better performance especially for weak signals than the method in, since the whole array, instead of the middle subarray in, is used for DOA estimation. Simulation results illustrate that both DOAs and mutual coupling coefficients can be estimated efficiently with the proposed method.

This paper proposes a new method for identification of time-varying autoregressive (TVAR) models based on local polynomial modeling (LPM) and applies it to investigate the dynamic spectral information of event-related electroencephalogram (EEG). The proposed method models the TVAR coefficients locally by polynomials and estimates those using least-squares estimation with a kernel having a certain bandwidth. A data-driven variable bandwidth selection method is developed to obtain the optimal bandwidth, which minimizes the mean squared error (MSE). Simulation results show that the LPM-based TVAR identification method outperforms conventional methods for different scenarios. The advantages of the LPM method make it a useful high-resolution time-frequency analysis (TFA) technique for nonstationary biomedical signals like EEG. Experimental results show that the LPM method can reveal more meaningful time-frequency characteristics than wavelet transform.

This paper studies the convergence behaviors of the noise-constrained normalized least mean squares (NCNLMS) algorithm recently proposed in the work of Chan et al. (2008). Like its LMS counterpart, the NCNLMS algorithm employs the prior knowledge of the additive noise to adjust its step-size. Following (Wei et al., 2001), the convergence behaviors of the NCLMS under the noise mismatch cases are firstly derived. Using a novel transformation approach and the small step-size properties of the NCNLMS algorithm at convergence, the mean and mean squares behaviors of this algorithm are derived. The validity of the proposed analysis is verified well by computer simulations and the relative merits of the NCLMS and NCNLMS algorithms are also compared.

This paper proposes a new two-stage method for restoring image corrupted by additive impulsive and Gaussian noise based on local polynomial regression (LPR) and edge preserving regularization. In LPR, the observations are modeled locally by a polynomial using least-squares criterion with a kernel controlled by a certain bandwidth matrix. A refined intersection confidence intervals (RICI) adaptive scale selector for symmetric kernel is applied in LPR to achieve a better bias-variance tradeoff. The method is further extended to steering kernel with local orientation to adapt better to local characteristics of images. The resulting steering-kernel-based LPR with RICI method (SK-LPR-RICI) is applied to smooth images contaminated with Gaussian noise. Furthermore, to remove the impulsive noise in images, an edge-preserving regularization method is employed prior to SK-LPR-RICI and it gives rise to a two-stage method for suppressing both additive impulsive and Gaussian noises. Simulation results show that the proposed method performs satisfactorily and the SK-LPR-RICI method significantly improves the performance after edge-preservation regularization in suppressing the impulsive noise.

This paper studies an M-estimation-based method for linear estimation with weighted L1 regularization and its recursive implementation. Motivated by the sensitivity of conventional least-squares-based L1-regularized linear estimation (Lasso) in impulsive noise environment, an M-estimator-based Lasso (M-Lasso) method is introduced to restrain the outliers and an iterative re-weighted least-squares (IRLS) algorithm is proposed to solve this M-estimation problem. Moreover, instead of using the matrix inversion formula, QR decomposition (QRD) is employed in the M-Lasso for recursive implementation with a lower arithmetic complexity. Simulation results show that the M-estimation-based Lasso performs considerably better than the traditional LS-based Lasso in suppressing the impulsive noise, and its recursive QRD algorithm has a good performance in online processing.

This paper introduces two new time-frequency analysis methods originated from the minimum variance spectral estimation (MVSE) for nonstationary time-series. First, a windowed MVSE (WMVSE) extends the conventional MVSE by windowing the observation data to obtain a time-frequency distribution for the time-series. Moreover, the window lengths are selected adaptively by the intersection of confidence intervals (ICI) rule to improve the time-frequency resolution. Secondly, a new recursive MVSE (RMVSE) is developed to process the input samples recursively at a lower arithmetic complexity for online time-frequency analysis. Simulation results show that the proposed WMVSE with adaptive windows offers better frequency resolutions than the Fourier-transformed-based time-frequency distributions, and the RMVSE has a good performance when tracking sinusoidal signals

This paper studies the time-frequency analysis of trunk response based on surface electromyography (EMG) under an experimental protocol of sudden load release. Due to difficulties in feature extraction of surface EMG in short-time muscular reflex using conventional time-frequency analysis algorithms, a novel method called adaptive Lomb periodogram (ALP) is employed to improve the time and frequency resolutions by adaptively selecting the window sizes. Experimental results demonstrate the improved time-frequency resolution of the ALP in analyzing the EMG signals compared with conventional methods using scalogram. The findings of this study would provide important information in devising objective and quantitative protocol for neuromuscular function assessment, which are applicable to rehabilitation of low back pain patients, motor control and training, elderly fall prevention and ergonomic studies.

This paper proposes a new minimum variance spectral estimation (MVSE)-based time-frequency analysis (TFA) for click-evoked otoacoustic emissions (CEOAEs). The conventional MVSE is extended to TFA by windowing the observation data to obtain a time-frequency distribution for the time-series. Based on the characteristics of CEOAEs, the window size is given a small value at high frequencies and a large value at low frequencies. The adaptive window size yields the proposed frequency-dependent WMVSE (FDWMVSE). The FDWMVSE integrates the advantages of adaptive window selection of wavelet transform and good resolution of MVSE. Experimental results show that the FDWMVSE can achieve better frequency resolution than other TFA methods when applied to synthesized and real CEOAEs

This paper presents a new Kalman filter-based subspace tracking algorithm and its application to directions of arrival (DOA) estimation. An autoregressive (AR) process is used to describe the dynamics of the subspace and a new adaptive Kalman filter with variable measurements (KFVM) algorithm is developed to estimate the time-varying subspace recursively from the state-space model and the given observations. For stationary subspace, the proposed algorithm will switch to the conventional PAST to lower the computational complexity. Simulation results show that the adaptive subspace tracking method has a better performance than conventional algorithms in DOA estimation for a wide variety of experimental condition

In this paper, we present a robust multiuser detector for wireless multicarrier code-division multiple access (MC-CDMA) systems under time-varying narrowband interference (NBI). The conventional least-squares (LS) channel estimators and multiuser detectors will perform poorly when narrowband interfering signals contaminate the multicarrier systems. A new weighted least M-estimate (WLM) multiuser detector is proposed to jointly suppress multiple access interference (MAI) and time-varying NBI. The WLM multiuser detector resorts to M-estimate and weighted least-squares (WLS) techniques. A weighted recursive least M-estimate (WRLM) channel estimator is exploited to estimate the time-varying frequency-selective fading channels in the presence of NBI. Numerical results show that the proposed WLM multiuser detector significantly outperforms over the conventional linear decorrelator, the robust decorrelating detector with M-estimate and the WLS detector under NBI

This paper proposes a new QR-decomposition-based recursive frequency estimation algorithm for multiple sinusoids based on the linear prediction (LP) approach. It extends the batch processing algorithm of So et al. in order to process the input samples recursively at a much lower arithmetic complexity for supporting on-line applications. Furthermore, a weighted least M-estimate (WLM) algorithm is developed to improve robustness to impulsive noise. Simulation results show that the robust recursive frequency estimator has a better performance than the conventional LS estimation in impulsive noise environment

This paper proposes a new Kalman filter-based algorithm for multichannel autoregressive (AR) spectrum estimation and adaptive coherence analysis with variable number of measurements. A stochastically perturbed k -order difference equation constraint model is used to describe the dynamics of the AR coefficients and the intersection of confidence intervals (ICI) rule is employed to determine the number of measurements adaptively to improve the time-frequency resolution of the AR spectrum and coherence function. Simulation results show that the proposed algorithm achieves a better time-frequency resolution than conventional algorithms for non-stationary signals

This paper presents a new Kalman filter-based power spectral density estimation (PSD) algorithm for nonstationary pressure signals. The pressure signal is assumed to be an autoregressive (AR) process, and a stochastically perturbed difference equation constraint model is used to describe the dynamics of the AR coefficients. The proposed Kalman filter frame uses variable number of measurements to estimate the time-varying AR coefficients and yield the PSD estimation with better time-frequency resolution. Simulation results show that the proposed algorithm achieves a better time-frequency resolution than conventional algorithms for nonstationary pressure signals

This article introduces a robust adaptive Lomb periodogram (RALP) for time-frequency (TF) analysis of a time series with sinusoidal and transient components, which are possibly non-uniformly sampled. It extends the conventional Lomb spectrum by windowing the observation data and adaptively selects the window lengths by the intersection of confidence intervals (ICI) rule. The influence of transient components on the conventional time-frequency representation can be moderated using M-estimation of robust statistics. Instead of treating the transient components as impulsive noise and removing them, the proposed RALP TF distribution yields separately a time domain representation of the transient components and a conventional TF representation of the sinusoidal components, which greatly improves the visualization and detection of these components. Simulation results show that the proposed RALP differentiates the two kinds of components well, and offers better time and frequency resolutions than the conventional Lomb periodogram.

The paper proposes a new method for performing multi-resolution analysis (MRA) of non-uniform data with jump discontinuities and impulsive noise using robust M-estimator-based local polynomial regression (LPR). The basic idea is to interpolate the smoothed estimate, after performing the robust LPR, on a uniform grid in order to perform the MRA using the ordinary wavelet transform. Simulation results show that the new approach performs better than traditional LS-based LPR in preserving jump discontinuities and suppressing isolated impulses when intersection confident intervals (ICI) bandwidth selection is employed.