Search

Recent usage of Virtual Reality (VR) technology in surgical training has emerged because of its cost-effectiveness, time savings, and cognition-based feedback generation. However, the quantitative evaluation of its effectiveness in training is still not studied thoroughly. This paper demonstrates the effectiveness of a VR-based surgical training simulator in laparoscopic surgery and investigates how stochastic modeling represented as Continuous-time Markov-chain (CTMC) can be used to explicit the training status of the surgeon. By comparing the training in real environments and in VR-based training simulators, the authors also explore the validity of the VR simulator in laparoscopic surgery. The study further aids in establishing learning models of surgeons, supporting continuous evaluation of training processes for the derivation of real-time feedback by CTMC-based modeling.

The limb position effect is a multi-faceted problem, associated with decreased upper-limb prosthesis control acuity following a change in arm position. Factors contributing to this problem can arise from distinct environmental or physiological sources. Despite their differences in origin, the effect of each factor manifests similarly as increased input data variability. This variability can cause incorrect decoding of user intent. Previous research has attempted to address this by better capturing input data variability with data abundance. In this paper, we take an alternative approach and investigate the effect of reducing trial-to-trial variability by improving the consistency of muscle activity through user training. Ten participants underwent 4 days of myoelectric training with either concurrent or delayed feedback in a single arm position. At the end of training participants experienced a zero-feedback retention test in multiple limb positions. In doing so, we tested how well the skill learned in a single limb position generalized to untrained positions. We found that delayed feedback training led to more consistent muscle activity across both the trained and untrained limb positions. Analysis of patterns of activations in the delayed feedback group suggest a structured change in muscle activity occurs across arm positions. Our results demonstrate that myoelectric user-training can lead to the retention of motor skills that bring about more robust decoding across untrained limb positions. This work highlights the importance of reducing motor variability with practice, prior to examining the underlying structure of muscle changes associated with limb position.

Magnetorheological (MR) fluid exhibits the ability to modulate its shear state through variations in magnetic field intensity, and is widely used for applications requiring damping. Traditional MR dampers use the current in the coil to adjust the magnetic field strength, but the accumulated heat can cause the magnetic field strength to decay if it works for a long time. In order to deal with this shortcoming, a novel MR damper is proposed in this paper, which is based on a variable displacement permanent magnet to adjust the output resistance torque and applied to an exoskeleton joint for human load transfer assistance. A finite element model is used to determine the size parameters of the magnet and separator, so that the maximum output torque is optimal and the torque is uniformly distributed with the magnet displacement. The MR damper was characterized and calibrated on the experimental bench to make it controllable. The novel design enables the torque mass density of the MR damper to reach 8.83Nmm/g, the torque volume density to reach 48.7N/mm2, and has stability for long-term operation. Based on the torque control method proposed, a preliminary human experiment is conducted. The ground reaction force (GRF) data of the subjects is analyzed here, which represents the effect of load transfer to the exoskeleton. Compared with no exoskeleton, the GRF with exoskeleton is significantly reduced: the peak GRF in early stance phase is reduced by 24.14%, and in late stance phase is reduced by 19.77%. Based on our net load benefit (NLB) and net force benefit (NFB) evaluation indicators, the effectiveness of the proposed MR damper exoskeleton for human weight bearing assistance is established.

With the goal of promoting the development of myoelectric control technology, this paper focuses on exploring graph neural network (GNN) based robust electromyography (EMG) pattern recognition solutions. Given that high-density surface EMG (HD-sEMG) signal contains rich temporal and spatial information, the multi-view spatial-temporal graph convolutional network (MSTGCN)is adopted as the basic classifier, and a feature extraction convolutional neural network (CNN) module is designed and integrated into MSTGCN to generate a new model called CNN-MSTGCN. The EMG pattern recognition experiments are conducted on HD-sEMG data of 17 gestures from 11 subjects. The ablation experiments show that each functional module of the proposed CNN-MSTGCN network has played a more or less positive role in improving the performance of EMG pattern recognition. The user-independent recognition experiments and the transfer learning-based cross-user recognition experiments verify the advantages of the proposed CNN-MSTGCN network in improving recognition rate and reducing user training burden. In the user-independent recognition experiments, CNN-MSTGCN achieves the recognition rate of 68%, which is significantly better than those obtained by residual network-50 (ResNet50, 47.5%, p < 0.001) and long-short-term-memory (LSTM, 57.1%, p=0.045). In the transfer learning-based cross-user recognition experiments, TL-CMSTGCN achieves an impressive recognition rate of 92.3%, which is significantly superior to both TL-ResNet50 (84.6%, p = 0.003) and TL-LSTM (85.3%, p = 0.008). The research results of this paper indicate that GNN has certain advantages in overcoming the impact of individual differences, and can be used to provide possible solutions for achieving robust EMG pattern recognition technology.

The existing surface electromyography-based pattern recognition system (sEMG-PRS) exhibits limited generalizability in practical applications. In this paper, we propose a stacked weighted random forest (SWRF) algorithm to enhance the long-term usability and user adaptability of sEMG-PRS. First, the weighted random forest (WRF) is proposed to address the issue of imbalanced performance in standard random forests (RF) caused by randomness in sampling and feature selection. Then, the stacking is employed to further enhance the generalizability of WRF. Specifically, RF is utilized as the base learner, while WRF serves as the meta-leaning layer algorithm. The SWRF is evaluated against classical classification algorithms in both online experiments and offline datasets. The offline experiments indicate that the SWRF achieves an average classification accuracy of 89.06%, outperforming RF, WRF, long short-term memory (LSTM), and support vector machine (SVM). The online experiments indicate that SWRF outperforms the aforementioned algorithms regarding long-term usability and user adaptability. We believe that our method has significant potential for practical application in sEMG-PRS.

Currently, resting-state electroencephalography (rs-EEG) has become an effective and low-cost evaluation way to identify autism spectrum disorders (ASD) in children. However, it is of great challenge to extract useful features from raw rs-EEG data to improve diagnosis performance. Traditional methods mainly rely on the design of manual feature extractors and classifiers, which are separately performed and cannot be optimized simultaneously. To this end, this paper proposes a new end-to-end diagnostic method based on a recently emerged graph convolutional neural network for the diagnosis of ASD in children. Inspired by related neuroscience findings on the abnormal brain functional connectivity and hemispheric asymmetry characteristics observed in autism patients, we design a new Regional-asymmetric Adaptive Graph Convolutional Neural Network (RAGNN). It utilizes a hierarchical feature extraction and fusion process to learn separable spatiotemporal EEG features from different brain regions, two hemispheres, and a global brain. In the temporal feature extraction section, we utilize a convolutional layer that spans from the brain area to the hemisphere. This allows for effectively capturing temporal features both within and between brain areas. To better capture spatial characteristics of multi-channel EEG signals, we employ adaptive graph convolutional learning to capture non-Euclidean features within the brain's hemispheres. Additionally, an attention layer is introduced to highlight different contributions of the left and right hemispheres, and the fused features are used for classification. We conducted a subject-independent cross-validation experiment on rs-EEG data from 45 children with ASD and 45 typically developing (TD) children. Experimental results have shown that our proposed RAGNN model outperformed several existing deep learning-based methods (ShaollowNet, EEGNet, TSception, ST-GCN, and CGRU-MDGN).

Walking is one of the most common daily movements of the human body. Therefore, quantitative evaluation of human walking has been commonly used to assist doctors in grasping the disease degree and rehabilitation process of patients in the clinic. Compared with the kinematic characteristics, the ground reaction force (GRF) during walking can directly reflect the dynamic characteristics of human walking. It can further help doctors understand the degree of muscle recovery and joint coordination of patients. This paper proposes a GRF estimation method based on the elastic elements and Newton-Euler equation hybrid driving GRF estimation method. Compared with the existing research, the innovations are as follows. 1) The hardware system consists of only two inertial measurement units (IMUs) placed on shanks. The acquisition of the overall motion characteristics of human walking is realized through the simplified four-link walking model and the thigh prediction method. 2) The method was validated not only on 10 healthy subjects but also on 11 Parkinson's patients and 10 stroke patients with normalized mean absolute errors (NMAEs) of 5.95%±1.32%, 6.09%±2.00%, 5.87%±1.59%. 3) This paper proposes a dynamic balance assessment method based on the acquired motion data and the estimated GRF. It evaluates the overall balance ability and fall risk at four key time points for all subjects recruited. Because of the low-cost system, ease of use, low motion interference and environmental constraints, and high estimation accuracy, the proposed GRF estimation method and walking balance automatic assessment have broad clinical value.

In the process of rehabilitation treatment for stroke patients, rehabilitation evaluation is a significant part in rehabilitation medicine. Researchers intellectualized the evaluation of rehabilitation evaluation methods and proposed quantitative evaluation methods based on evaluation scales, without the clinical background of physiatrist. However, in clinical practice, the experience of physiatrist plays an important role in the rehabilitation evaluation of patients. Therefore, this paper designs a 5 degrees of freedom (DoFs) upper limb (UL) rehabilitation robot and proposes a rehabilitation evaluation model based on Belief Rule Base (BRB) which can add the expert knowledge of physiatrist to the rehabilitation evaluation. The motion data of stroke patients during active training are collected by the rehabilitation robot and signal collection system, and then the upper limb motor function of the patients is evaluated by the rehabilitation evaluation model. To verify the accuracy of the proposed method, Back Propagation Neural Network (BPNN) and Support Vector Machines (SVM) are used to evaluate. Comparative analysis shows that the BRB model has high accuracy and effectiveness among the three evaluation models. The results show that the rehabilitation evaluation model of stroke patients based on BRB could help physiatrists to evaluate the UL motor function of patients and master the rehabilitation status of stroke patients.

Automatic identification of visual learning style in real time using raw electroencephalogram (EEG) is challenging. In this work, inspired by the powerful abilities of deep learning techniques, deep learning-based models are proposed to learn high-level feature representation for EEG visual learning identification. Existing computer-aided systems that use electroencephalograms and machine learning can reasonably assess learning styles. Despite their potential, offline processing is often necessary to eliminate artifacts and extract features, making these methods unsuitable for real-time applications. The dataset was chosen with 34 healthy subjects to measure their EEG signals during resting states (eyes open and eyes closed) and while performing learning tasks. The subjects displayed no prior knowledge of the animated educational content presented in video format. The paper presents an analysis of EEG signals measured during a resting state with closed eyes using three deep learning techniques: Long-term, short-term memory (LSTM), Long-term, short-term memory-convolutional neural network (LSTM-CNN), and Long-term, short-term memory-Fully convolutional neural network (LSTM-FCNN). The chosen techniques were based on their suitability for real-time applications with varying data lengths and the need for less computational time. The optimization of hypertuning parameters has enabled the identification of visual learners through the implementation of three techniques. LSTM-CNN technique has the highest average accuracy of 94%, a sensitivity of 80%, a specificity of 92%, and an F1 score of 94% when identifying the visual learning style of the student out of all three techniques. This research has shown that the most effective method is the deep learning-based LSTM-CNN technique, which accurately identifies a student's visual learning style.

Motor imagery (MI) decoding methods are pivotal in advancing rehabilitation and motor control research. Effective extraction of spectral-spatial-temporal features is crucial for MI decoding from limited and low signal-to-noise ratio electroencephalogram (EEG) signal samples based on brain-computer interface (BCI). In this paper, we propose a lightweight Multi-Feature Attention Neural Network (M-FANet) for feature extraction and selection of multi-feature data. M-FANet employs several unique attention modules to eliminate redundant information in the frequency domain, enhance local spatial feature extraction and calibrate feature maps. We introduce a training method called Regularized Dropout (R-Drop) to address training-inference inconsistency caused by dropout and improve the model's generalization capability. We conduct extensive experiments on the BCI Competition IV 2a (BCIC-IV-2a) dataset and the 2019 World robot conference contest-BCI Robot Contest MI (WBCIC-MI) dataset. M-FANet achieves superior performance compared to state-of-the-art MI decoding methods, with 79.28% 4-class classification accuracy (kappa: 0.7259) on the BCIC-IV-2a dataset and 77.86% 3-class classification accuracy (kappa: 0.6650) on the WBCIC-MI dataset. The application of multi-feature attention modules and R-Drop in our lightweight model significantly enhances its performance, validated through comprehensive ablation experiments and visualizations.

This paper leverages the OpenSim physics-based simulation environment for the forward dynamic simulation of an osseointegrated transfemoral amputee musculoskeletal model, wearing a generic prosthesis. A deep reinforcement learning architecture, which combines the proximal policy optimization algorithm with imitation learning, is designed to enable the model to walk by using three different observation states. The first is a complete state that includes the agent's kinematics, ground reaction forces, and muscle data; the second is a reduced state that only includes the kinematics and ground reaction forces; the third is an augmented state that combines the kinematics and ground reaction forces with a prediction of the muscle data generated by a fully-connected feed-forward neural network. The empirical results demonstrate that the model trained with the augmented observation state can achieve walking patterns with rewards and gait symmetry ratings comparable to those of the model trained with the complete observation state, while there are no symmetric walking patterns when using the reduced observation state. This paper shows the importance of including muscle data in a deep reinforcement learning architecture for the forward dynamic simulation of musculoskeletal models of transfemoral amputees.

The self-aligning capability of an exoskeleton is important to ensure wearing comfort, and the delicate motion ability of the exoskeleton is essential for motion assistance. Designing a self-aligning exoskeleton that offers improved wearing comfort and enhanced motion-assistance functions remains a challenge. This paper proposes a novel spatial self-aligning mechanism for a knee exoskeleton to enable simultaneous assistance in the flexion and extension (FE) of the knee joint and the internal and external rotation (IER) of the hip joint. Additionally, considering the misalignment of the human-robot joint axes, a kinematic model of the knee exoskeleton is established and analyzed to demonstrate the kinematic compatibility of the exoskeleton. Furthermore, a global torque manipulability (GTM) index is proposed to evaluate the effects of dimensional parameters on the exoskeleton's performance, and then the knee exoskeleton is optimized according to the GTM index. Finally, experiments are conducted to validate the performance of the proposed exoskeleton. The experimental results show that during knee FE and hip IER, the proposed exoskeleton exhibits lower interaction forces and torques than existing exoskeletons.

Electroencephalography (EEG) signals are the brain signals acquired using the non-invasive approach. Owing to the high portability and practicality, EEG signals have found extensive application in monitoring human physiological states across various domains. In recent years, deep learning methodologies have been explored to decode the intricate information embedded in EEG signals. However, since EEG signals are acquired from humans, it has issues with acquiring enormous amounts of data for training the deep learning models. Therefore, previous research has attempted to develop pre-trained models that could show significant performance improvement through fine-tuning when data are scarce. Nonetheless, existing pre-trained models often struggle with constraints, such as the necessity to operate within datasets of identical configurations or the need to distort the original data to apply the pre-trained model. In this paper, we proposed the domain-free transformer, called DFformer, for generalizing the EEG pre-trained model. In addition, we presented the pre-trained model based on DFformer, which is capable of seamless integration across diverse datasets without necessitating architectural modification or data distortion. The proposed model achieved competitive performance across motor imagery and sleep stage classification datasets. Notably, even when fine-tuned on datasets distinct from the pre-training phase, DFformer demonstrated marked performance enhancements. Hence, we demonstrate the potential of DFformer to overcome the conventional limitations in pre-trained model development, offering robust applicability across a spectrum of domains.

Motor imagery (MI) based brain-computer interfaces (BCIs) enable the direct control of external devices through the imagined movements of various body parts. Unlike previous systems that used fixed-length EEG trials for MI decoding, asynchronous BCIs aim to detect the user's MI without explicit triggers. They are challenging to implement, because the algorithm needs to first distinguish between resting-states and MI trials, and then classify the MI trials into the correct task, all without any triggers. This paper proposes a sliding window prescreening and classification (SWPC) approach for MI-based asynchronous BCIs, which consists of two modules: a prescreening module to screen MI trials out of the resting-state, and a classification module for MI classification. Both modules are trained with supervised learning followed by self-supervised learning, which refines the feature extractors. Within-subject and cross-subject asynchronous MI classifications on four different EEG datasets validated the effectiveness of SWPC, i.e., it always achieved the highest average classification accuracy, and outperformed the best state-of-the-art baseline on each dataset by about 2%.

This study presents a novel method to assess the learning effectiveness using Electroencephalography (EEG)-based deep learning model. It is difficult to assess the learning effectiveness of professional courses in cultivating students' ability objectively by questionnaire or other assessment methods. Research in the field of brain has shown that innovation ability can be reflected from cognitive ability which can be embodied by EEG signal features. Three navigation tasks of increasing cognitive difficulty were designed and a total of 41 subjects participated in the experiment. For the classification and tracking of the subjects' EEG signals, a convolutional neural network (CNN)-based Multi-Time Scale Spatiotemporal Compound Model (MTSC) is proposed in this paper to extract and classify the features of the subjects' EEG signals. Furthermore, Spiking neural networks (SNN) -based NeuCube is used to assess the learning effectiveness and demonstrate cognitive processes, acknowledging that NeuCube is an excellent method to display the spatiotemporal differences between spikes emitted by neurons. The results of the classification experiment show that the cognitive training traces of different students in solving three navigational problems can be effectively distinguished. More importantly, new information about navigation is revealed through the analysis of feature vector visualization and model dynamics. This work provides a foundation for future research on cognitive navigation and the training of students' navigational skills.

Predicting individual behavior from brain imaging data using machine learning is a rapidly growing field in neuroscience. Functional connectivity (FC), which captures interactions between different brain regions, contains valuable information about the organization of the brain and is considered a crucial feature for modeling human behavior. Graph convolutional networks (GCN) have proven to be a powerful tool for extracting graph structure features and have shown promising results in various FC-based classification tasks, such as disease classification and prognosis prediction. Despite this success, few behavior prediction models currently exist based on GCN, and their performance is not satisfactory. To address this gap, a new model called the Multi-Scale FC-based Multi-Order GCN (MSFC-MO-GCN) was proposed in this paper. The model considers the hierarchical structure of the brain system and utilizes FCs inferred from multiple spatial scales as input to comprehensively characterize individual brain organization. To enhance the feature learning ability of GCN, a multi-order graph convolutional layer is incorporated, which uses multi-order neighbors to guide message passing and learns high-order graph information of nodal connections. Additionally, an inter-subject contrast constraint is designed to control the potential information redundancy of FCs among different spatial scales during the feature learning process. Experimental evaluation were conducted on the publicly available dataset from human connectome project. A total of 805 healthy subjects were included and 5 representative behavior metrics were used. The experimental results show that our proposed method outperforms the existing behavior prediction models in all behavior prediction tasks.

Gait coordination (GC), meaning that one leg moves in the same pattern but with a specific phase lag to the other, is a spontaneous behavior in the walking of a healthy person. It is also crucial for unilateral amputees with the robotic leg prosthesis to perform ambulation cooperatively in the real world. However, achieving the GC for amputees poses significant challenges to the prostheses' dynamic modeling and control design. Still, there has not been a clear survey on the initiation and evolution of the detailed solutions, hindering the precise decision of future explorations. To this end, this paper comprehensively reviews GC-oriented dynamic modeling and adaptive control methods for robotic leg prostheses. Considering the two representative environments concerned with adaptive control, we first classify the dynamic models into the deterministic model for structured terrain and the constrained stochastic model for stochastically uneven terrain. Inspired by the concept of synchronization, we then emphasize three typical problems for the GC realization, i.e., complete coordination on structured terrain, stochastic coordination on stochastically uneven terrain, and finite-time delayed stochastic coordination. Finally, we conclude with a discussion on the remaining challenges and opportunities in controlling robotic leg prostheses.

Electroencephalography (EEG) datasets are characterized by low signal-to-noise signals and unquantifiable noisy labels, which hinder the classification performance in rapid serial visual presentation (RSVP) tasks. Previous approaches primarily relied on supervised learning (SL), which may result in overfitting and reduced generalization performance. In this paper, we propose a novel multi-task collaborative network (MTCN) that integrates both SL and self-supervised learning (SSL) to extract more generalized EEG representations. The original SL task, i.e., the RSVP EEG classification task, is used to capture initial representations and establish classification thresholds for targets and non-targets. Two SSL tasks, including the masked temporal/spatial recognition task, are designed to enhance temporal dynamics extraction and capture the inherent spatial relationships among brain regions, respectively. The MTCN simultaneously learns from multiple tasks to derive a comprehensive representation that captures the essence of all tasks, thus mitigating the risk of overfitting and enhancing generalization performance. Moreover, to facilitate collaboration between SL and SSL, MTCN explicitly decomposes features into task-specific features and task-shared features, leveraging both label information with SL and feature information with SSL. Experiments conducted on THU, CAS, and GIST datasets illustrate the significant advantages of learning more generalized features in RSVP tasks. Our code is publicly accessible at https://github.com/Tammie-Li/MTCN.

Alzheimer's Disease (AD) is a widespread, chronic, irreversible, and degenerative condition, and its early detection during the prodromal stage is of utmost importance. Typically, AD studies rely on single data modalities, such as MRI or PET, for making predictions. Nevertheless, combining metabolic and structural data can offer a comprehensive perspective on AD staging analysis. To address this goal, this paper introduces an innovative multi-modal fusion-based approach named as Dual-3DM3-AD. This model is proposed for an accurate and early Alzheimer's diagnosis by considering both MRI and PET image scans. Initially, we pre-process both images in terms of noise reduction, skull stripping and 3D image conversion using Quaternion Non-local Means Denoising Algorithm (QNLM), Morphology function and Block Divider Model (BDM), respectively, which enhances the image quality. Furthermore, we have adapted Mixed-transformer with Furthered U-Net for performing semantic segmentation and minimizing complexity. Dual-3DM3-AD model is consisted of multi-scale feature extraction module for extracting appropriate features from both segmented images. The extracted features are then aggregated using Densely Connected Feature Aggregator Module (DCFAM) to utilize both features. Finally, a multi-head attention mechanism is adapted for feature dimensionality reduction, and then the softmax layer is applied for multi-class Alzheimer's diagnosis. The proposed Dual-3DM3-AD model is compared with several baseline approaches with the help of several performance metrics. The final results unveil that the proposed work achieves 98% of accuracy, 97.8% of sensitivity, 97.5% of specificity, 98.2% of f-measure, and better ROC curves, which outperforms other existing models in multi-class Alzheimer's diagnosis.

Major Depression Disorder (MDD) is a common yet destructive mental disorder that affects millions of people worldwide. Making early and accurate diagnosis of it is very meaningful. Recently, EEG, a non-invasive technique of recording spontaneous electrical activity of brains, has been widely used for MDD diagnosis. However, there are still some challenges in data quality and data size of EEG: (1) A large amount of noise is inevitable during EEG collection, making it difficult to extract discriminative features from raw EEG; (2) It is difficult to recruit a large number of subjects to collect sufficient and diverse data for model training. Both of the challenges cause the overfitting problem, especially for deep learning methods. In this paper, we propose DiffMDD, a diffusion-based deep learning framework for MDD diagnosis using EEG. Specifically, we extract more noise-irrelevant features to improve the model's robustness by designing the Forward Diffusion Noisy Training Module. Then we increase the size and diversity of data to help the model learn more generalized features by designing the Reverse Diffusion Data Augmentation Module. Finally, we re-train the classifier on the augmented dataset for MDD diagnosis. We conducted comprehensive experiments to test the overall performance and each module's effectiveness. The framework was validated on two public MDD diagnosis datasets, achieving the state-of-the-art performance.

Brain-computer interfaces (BCIs) can enable direct communication with assistive devices by recording and decoding signals from the brain. To achieve high performance, many electrodes will be used, such as the recently developed invasive BCIs with channel numbers up to hundreds or even thousands. For those high-throughput BCIs, channel selection is important to reduce signal redundancy and invasiveness while maintaining decoding performance. However, such endeavour is rarely reported for invasive BCIs, especially those using deep learning methods. Two deep learning-based methods, referred to as Gumbel and STG, were proposed in this paper. They were evaluated using the Stereo-electroencephalography (SEEG) signals, and compared with three other methods, including manual selection, mutual information-based method (MI), and all channels (all channels without selection). The task is to classify the SEEG signals into five movements using channels selected by each method. When 10 channels were selected, the mean classification accuracies using Gumbel, STG (referred to as STG-10), manual selection, and MI selection were 65%, 60%, 60%, and 47%, respectively, whilst the accuracy was 59% using all channels (no selection). In addition, an investigation of the selected channels showed that Gumbel and STG have successfully identified the pre-central and post-central areas, which are closely related to motor control. Both Gumbel and STG successfully selected the informative channels in SEEG recordings while maintaining decoding accuracy. This study enables future high-throughput BCIs using deep learning methods, to identify useful channels and reduce computing and wireless transmission pressure.

Cross-subject electroencephalogram (EEG) based seizure subtype classification is very important in precise epilepsy diagnostics. Deep learning is a promising solution, due to its ability to automatically extract latent patterns. However, it usually requires a large amount of training data, which may not always be available in clinical practice. This paper proposes Multi-Branch Mutual-Distillation (MBMD) Transformer for cross-subject EEG-based seizure subtype classification, which can be effectively trained from small labeled data. MBMD Transformer replaces all even-numbered encoder blocks of the vanilla Vision Transformer by our designed multi-branch encoder blocks. A mutual-distillation strategy is proposed to transfer knowledge between the raw EEG data and its wavelets of different frequency bands. Experiments on two public EEG datasets demonstrated that our proposed MBMD Transformer outperformed several traditional machine learning and state-of-the-art deep learning approaches. To our knowledge, this is the first work on knowledge distillation for EEG-based seizure subtype classification.

Sleep abnormalities can have severe health consequences. Automated sleep staging, i.e. labelling the sequence of sleep stages from the patient's physiological recordings, could simplify the diagnostic process. Previous work on automated sleep staging has achieved great results, mainly relying on the EEG signal. However, often multiple sources of information are available beyond EEG. This can be particularly beneficial when the EEG recordings are noisy or even missing completely. In this paper, we propose CoRe-Sleep, a Coordinated Representation multimodal fusion network that is particularly focused on improving the robustness of signal analysis on imperfect data. We demonstrate how appropriately handling multimodal information can be the key to achieving such robustness. CoRe-Sleep tolerates noisy or missing modalities segments, allowing training on incomplete data. Additionally, it shows state-of-the-art performance when testing on both multimodal and unimodal data using a single model on SHHS-1, the largest publicly available study that includes sleep stage labels. The results indicate that training the model on multimodal data does positively influence performance when tested on unimodal data. This work aims at bridging the gap between automated analysis tools and their clinical utility.

Autism spectrum disorder (ASD) one of the fastest-growing diseases in the world is a group of neurodevelopmental disorders. Eye movement as a biomarker and clinical manifestation represents unconscious brain processes that can objectively disclose abnormal eye fixation of ASD. With the aid of eye-tracking technology, plentiful methods that identify ASD based on eye movements have been developed, but there are rarely works specifically for scanpaths. Scanpaths as visual representations describe eye movement dynamics on stimuli. In this paper, we propose a scanpath-based ASD detection method, which aims to learn the atypical visual pattern of ASD through continuous dynamic changes in gaze distribution. We extract four sequence features from scanpaths that represent changes and the differences in feature space and gaze behavior patterns between ASD and typical development (TD) are explored based on two similarity measures, multimatch and dynamic time warping (DTW). It indicates that ASD children show more individual specificity, while normal children tend to develop similar visual patterns. The most noticeable contrasts lie in the duration of attention and the spatial distribution of visual attention along the vertical direction. Classification is performed using Long Short-Term Memory (LSTM) network with different structures and variants. The experimental results show that LSTM network outperforms traditional machine learning methods.

Deep brain stimulation (DBS) is establishing itself as a promising treatment for disorders of consciousness (DOC). Measuring consciousness changes is crucial in the optimization of DBS therapy for DOC patients. However, conventional measures use subjective metrics that limit the investigations of treatment-induced neural improvements. The focus of this study is to analyze the regulatory effects of DBS and explain the regulatory mechanism at the brain functional level for DOC patients. Specifically, this paper proposed a dynamic brain temporal-spectral analysis method to quantify DBS-induced brain functional variations in DOC patients. Functional near-infrared spectroscopy (fNIRS) that promised to evaluate consciousness levels was used to monitor brain variations of DOC patients. Specifically, a fNIRS-based experimental procedure with auditory stimuli was developed, and the brain activities during the procedure from thirteen DOC patients before and after the DBS treatment were recorded. Then, dynamic brain functional networks were formulated with a sliding-window correlation analysis of phase lag index. Afterwards, with respect to the temporal variations of global and regional networks, the variability of global efficiency, local efficiency, and clustering coefficient were extracted. Further, dynamic networks were converted into spectral representations by graph Fourier transform, and graph energy and diversity were formulated to assess the spectral global and regional variability. The results showed that DOC patients under DBS treatment exhibited increased global and regional functional variability that was significantly associated with consciousness improvements. Moreover, the functional variability in the right brain regions had a stronger correlation with consciousness enhancements than that in the left brain regions. Therefore, the proposed method well signifies DBS-induced brain functional variations in DOC patients, and the functional variability may serve as promising biomarkers for consciousness evaluations in DOC patients.

The traditional way of reading through Braille books is constraining the reading experience of blind or visually impaired (BVI) in the digital age. In order to improve the reading convenience of BVI, this paper proposes a low-cost and refreshable Braille display device, and solves the problems of high energy consumption and low latching force existing in existing devices. Further, the Braille display device was combined with the 3D Systems Touch device to develop an active Braille touch-reading system for digital reading of BVI with the help of the CHAI3D virtual environment. Firstly, according to the actual needs of BVI to touch and read the Braille dots, this paper utilizes the beam structure to provide a full latching function for the raised Braille dot without energy consumption. Through theoretical derivation and finite element analysis, the performance of the Braille dot actuator is optimized to provide sufficient feedback force and latching force for finger's touch-reading. Then, this paper designs a virtual Braille interactive environment based on the CHAI3D, and combines the sense of touch with audio to effectively improve the recognition accuracy and reading efficiency of BVI for Braille through the multi-modal presentation of Braille information. The performance test results of the device show that the average lifting force of the Braille dot actuator is 101.67 mN, the latching force is over 5 N, and the average refresh frequency is 17.1 Hz, which meets the touch-reading needs of BVI. User experiments show that the average accuracy rate of BVI subjects in identifying digitized Braille is 95.5%, and subjects have a high subjective evaluation of the system.

Individuals who have suffered a spinal cord injury often require assistance to complete daily activities, and for individuals with tetraplegia, recovery of upper-limb function is among their top priorities. Hybrid functional electrical stimulation (FES) and exoskeleton systems have emerged as a potential solution to provide upper limb movement assistance. These systems leverage the user's own muscles via FES and provide additional movement support via an assistive exoskeleton. To date, these systems have focused on single joint movements, limiting their utility for the complex movements necessary for independence. In this paper, we extend our prior work on model predictive control (MPC) of hybrid FES-exo systems and present a multi degree of freedom (DOF) hybrid controller that uses the controller's cost function to achieve desired behavior. In studies with neurologically intact individuals, the hybrid controller is compared to an exoskeleton acting alone for movement assistance scenarios incorporating multiple degrees-of-freedom of the limb to explore the potential for exoskeleton power consumption reduction and impacts on tracking accuracy. Additionally, each scenario is explored in simulation using the models required to generate the MPC formulation. The two DOF hybrid controller implementation saw reductions in power consumption and satisfactory trajectory tracking in both the physical and simulated systems. In the four DOF implementation, the experimental results showed minor improvements for some joints of the upper limb. In simulation, we observed comparable performance as in the two DOF implementation.

Recent developments in dexterous myoelectric prosthetics have established a hardware base for human-machine interfaces. Although pattern recognition techniques have seen successful deployment in gesture classification, their applications remain largely confined to certain specific discrete gestures. Addressing complex daily tasks demands an immediate need for precise simultaneous and proportional control (SPC) for multiple degrees of freedom (DoFs) movements. In this paper, we introduce an SPC approach for multi-DoF wrist movements using the cumulative spike trains (CSTs) of motor unit pools, merely leveraging single-DoF training. The efficacy of our proposed approach was validated offline against existing methods respectively based on non-negative matrix factorization and motor unit spike trains, using experimental data. The experimental process includes both single-DoF (for training) and multi-DoF (for testing) movements. We evaluated the performance using Pearson correlation coefficient (R) and the normalized root mean square error (nRMSE). The results reveal that our method outperforms comparative approaches in force estimation for both testing datasets (3 and 4). On average, for dataset 3, R and nRMSE of the flexion/extension DoF (the pronation/supination DoF) are 0.923±0.037 (0.901±0.040) and 12.3±3.1% (12.9±2.2%); similarly, those of dataset 4 are 0.865±0.057 (0.837±0.053) and 14.9±2.9% (15.4±2.0%), respectively. The outcomes demonstrate the effectiveness of our method in simultaneous and proportional force estimation for multi-DoF wrist movements, showing a promising potential as a neural-machine interface for SPC of dexterous myoelectric prostheses.

During the postoperative hand rehabilitation period, it is recommended that the repaired flexor tendons be continuously glided with sufficient tendon excursion and carefully managed protection to prevent adhesion with adjacent tissues. Thus, finger joints should be passively mobilized through a wide range of motion (ROM) with physiotherapy. During passive mobilization, sequential flexion of the metacarpophalangeal (MCP) joint followed by the proximal interphalangeal (PIP) joint is recommended for maximizing tendon excursion. This paper presents a lightweight device for postoperative flexor tendon rehabilitation that uses a single motor to achieve sequential joint flexion movement. The device consists of an orthosis, a cable, and a single motor. The degree of spatial stiffness and cable path of the orthosis were designed to apply a flexion moment to the MCP joint prior to the PIP joint. The device was tested on both healthy individuals and a patient who had undergone flexor tendon repair surgery, and both flexion and extension movement could be achieved with a wide ROM and sequential joint flexion movement using a single motor.

The coupled analysis of corticomuscular function based on physiological electrical signals can identify differences in causal relationships between electroencephalogram (EEG) and surface electromyogram (sEMG) in different motor states. The existing methods are mainly devoted to the analysis in the same frequency band, while ignoring the cross-band coupling, which plays an active role in motion control. Considering the inherent multiscale characteristics of physiological signals, a method combining Ordinal Partition Transition Networks (OPTNs) and Multivariate Variational Modal Decomposition (MVMD) was proposed in this paper. The EEG and sEMG were firstly decomposed on a time-frequency scale using MVMD, and then the coupling strength was calculated by the OPTNs to construct a corticomuscular coupling network, which was analyzed with complex network parameters. Experimental data were obtained from a self-acquired dataset consisting of EEG and sEMG of 16 healthy subjects at different sizes of constant grip force. The results showed that the method was superior in representing changes in the causal link among multichannel signals characterized by different frequency bands and grip strength patterns. Complex information transfer between the cerebral cortex and the corresponding muscle groups during constant grip force output from the human upper limb. Furthermore, the sEMG of the flexor digitorum superficialis (FDS) in the low frequency band is the hub in the effective information transmission between the cortex and the muscle, while the importance of each frequency component in this transmission network becomes more dispersed as the grip strength grows, and the increase in coupling strength and node status is mainly in the γ band (30~60Hz). This study provides new ideas for deconstructing the mechanisms of neural control of muscle movements.

Modeling the muscle response to functional electrical stimulation (FES) is an important step during model-based FES control system design. The Hammerstein structure is widely used in simulating this nonlinear biomechanical response. However, a fixed relationship cannot cope well with the time-varying property of muscles and muscle fatigue. In this paper, we proposed an adaptive Hammerstein model to predict ankle joint torque induced by electrical stimulation, which used variable forgetting factor recursive least squares (VFFRLS) method to update the model parameters. To validate the proposed model, ten healthy individuals were recruited for short-duration FES experiments, ten for long-duration FES experiments, and three stroke patients for both. The isometric ankle dorsiflexion torque induced by FES was measured, and then the test performance of the fixed-parameter Hammerstein model, the adaptive Hammerstein model based on fixed forgetting factor recursive least squares (FFFRLS) and the adaptive Hammerstein model based on VFFRLS was compared. The goodness of fit, root mean square error, peak error and success rate were applied to evaluate the accuracy and stability of the model. The results indicate a significant improvement in both the accuracy and stability of the proposed adaptive model compared to the fixed-parameter model and the adaptive model based on FFFRLS. The proposed adaptive model enhances the ability of the model to cope with muscle changes.

Many rehabilitation exoskeletons have been used in the field of stroke rehabilitation. Generating human-like motion is necessary for exoskeletons to help patients perform activities of daily living (ADL) while maintaining interaction quality and ergonomics. However, most of the current motion generation algorithms utilize inverse kinematics (IK) to solve the final configuration before generation, and do not consider the movement of shoulder girdle. Separately considering the shoulder girdle motion and arm motion, this paper proposes an algorithm integrated IK to generate human-like motion. The arm moves towards the target with a bell-shaped velocity in the absence of the final configuration, and the shoulder girdle maintain natural passive motion. Moreover, the generated motion can be mapped to the configuration space of exoskeletons. Compared with the experimental data collected using a motion capture system, the values of RMSE and HPDI of the generated wrist trajectory in the task space are within 0.2 and 0.17, respectively, while those of RMSE in the joint space are within 15 deg, which demonstrates the human-like nature of the generated motion.

This paper introduces a walking path generation method specifically developed for the Smart Cane, which is a RNA (Robotic Navigation Assistance Device) aimed at enhancing indoor navigation for visually impaired individuals. The proposed approach combines the utilization of a LIPM (Linear Inverse Pendulum Model) and LFPC (Linear Foot Placement Controller) motion primitives to generate walking paths specifically designed for visually impaired individuals. The primary objective is to generate paths that conform to human motion constraints, thereby guaranteeing an efficient and natural navigation experience. Integrating autonomous navigation framework, the Smart Cane facilitates safe and effective guidance for visually impaired participants in the indoor environments. Furthermore, comparative experiments have been conducted to validate the effectiveness of the proposed method, providing evidence of its capability to generate walking paths that conform to human motion constraints. The experiment results indicate that the proposed walking path generation method is a promising solution to enhance the navigation experience of visually impaired individuals.

Falls represent a significant cause of injury among the elderly population. Extensive research has been devoted to the utilization of wearable IMU sensors in conjunction with machine learning techniques for fall detection. To address the challenge of acquiring costly training data, this paper presents a novel method that generates a substantial volume of synthetic IMU data with minimal actual fall experiments. First, unmarked 3D motion capture technology is employed to reconstruct human movements. Subsequently, utilizing the biomechanical simulation platform Opensim and forward kinematic methods, an ample amount of training data from various body segments can be custom generated. Synthetic IMU data was then used to train a machine learning model, achieving testing accuracies of 91.99% and 86.62% on two distinct datasets of actual fall-related IMU data. Building upon the simulation framework, this paper further optimized the single IMU attachment position and multiple IMU combinations on fall detection. The proposed method simplifies fall detection data acquisition experiments, provides novel venue for generating low cost synthetic data in scenario where acquiring data for machine learning is challenging and paves the way for customizing machine learning configurations.

Computational biomechanical analysis plays a pivotal role in understanding and improving human movements and physical functions. Although physics-based modeling methods can interpret the dynamic interaction between the neural drive to muscle dynamics and joint kinematics, they suffer from high computational latency. In recent years, data-driven methods have emerged as a promising alternative due to their fast execution speed, but label information is still required during training, which is not easy to acquire in practice. To tackle these issues, this paper presents a novel physics-informed deep learning method to predict muscle forces without any label information during model training. In addition, the proposed method could also identify personalized muscle-tendon parameters. To achieve this, the Hill muscle model-based forward dynamics is embedded into the deep neural network as the additional loss to further regulate the behavior of the deep neural network. Experimental validations on the wrist joint from six healthy subjects are performed, and a fully connected neural network (FNN) is selected to implement the proposed method. The predicted results of muscle forces show comparable or even lower root mean square error (RMSE) and higher coefficient of determination compared with baseline methods, which have to use the labeled surface electromyography (sEMG) signals, and it can also identify muscle-tendon parameters accurately, demonstrating the effectiveness of the proposed physics-informed deep learning method.

Nowadays, numerous countries are facing the challenge of aging population. Additionally, the number of people with reduced mobility due to physical illness is increasing. In response to this issue, robots used for walking assistance and sit-to-stand (STS) transition have been introduced in nursing to assist these individuals with walking. Given the shared characteristics of these robots, this paper collectively refers to them as Walking Support Robots (WSR). Additionally, service robots with assisting functions have been included in the scope of this review. WSR are a crucial element of modern nursing assistants and have received significant research attention. Unlike passive walkers that require much user's strength to move, WSR can autonomously perceive the state of the user and environment, and select appropriate control strategies to assist the user in maintaining balance and movement. This paper offers a comprehensive review of recent literature on WSR, encompassing an analysis of structure design, perception methods, control strategies and safety & comfort features. In conclusion, it summarizes the key findings, current challenges and discusses potential future research directions in this field.

The magnetoencephalogram (MEG) based on array optically pumped magnetometers (OPMs) has the potential of replacing conventional cryogenic superconducting quantum interference device. Phase synchronization is a common method for measuring brain oscillations and functional connectivity. Verifying the feasibility and fidelity of OPM-MEG in measuring phase synchronization will help its widespread application in the study of aforementioned neural mechanisms. The analysis method on source-level time series can weaken the influence of instantaneous field spread effect. In this paper, the OPM-MEG was used for measuring the evoked responses of 20Hz rhythmic and arrhythmic median nerve stimulation, and the inter-trial phase synchronization (ITPS) and inter-reginal phase synchronization (IRPS) of primary somatosensory cortex (SI) and secondary somatosensory cortex (SII) were analysed. The results find that under rhythmic condition, the evoked responses of SI and SII show continuous oscillations and the effect of resetting phase. The values of ITPS and IRPS significantly increase at the stimulation frequency of 20Hz and its harmonic of 40Hz, whereas the arrhythmic stimulation does not exhibit this phenomenon. Moreover, in the initial stage of stimulation, the ITPS and IRPS values are significantly higher at Mu rhythm in the rhythmic condition compared to arrhythmic. In conclusion, the results demonstrate the ability of OPM-MEG in measuring phase pattern and functional connectivity on source-level, and may also prove beneficial for the study on the mechanism of rhythmic stimulation therapy for rehabilitation.

Alzheimer's Disease (AD) accounts for the majority of dementia, and Mild Cognitive Impairment (MCI) is the early stage of AD. Early and accurate diagnosis of dementia plays a vital role in more targeted treatments and effectively halting disease progression. However, the clinical diagnosis of dementia requires various examinations, which are expensive and require a high level of expertise from the doctor. In this paper, we proposed a classification method based on multi-modal data including Electroencephalogram (EEG), eye tracking and behavioral data for early diagnosis of AD and MCI. Paradigms with various task difficulties were used to identify different severity of dementia: eye movement task and resting-state EEG tasks were used to detect AD, while eye movement task and delayed match-to-sample task were used to detect MCI. Besides, the effects of different features were compared and suitable EEG channels were selected for the detection. Furthermore, we proposed a data augmentation method to enlarge the dataset, designed an extra ERPNet feature extract layer to extract multi-modal features and used domain-adversarial neural network to improve the performance of MCI diagnosis. We achieved an average accuracy of 88.81% for MCI diagnosis and 100% for AD diagnosis. The results of this paper suggest that our classification method can provide a feasible and affordable way to diagnose dementia.

Upper limb functional impairments persisting after stroke significantly affect patients' quality of life. Precise adjustment of robotic assistance levels based on patients' motion intentions using sEMG signals is crucial for active rehabilitation. This paper systematically reviews studies on continuous prediction of upper limb single joints and multi-joint combinations motion intention using Model-Based (MB) and Model-Free (MF) approaches over the past decade, based on 186 relevant studies screened from six major electronic databases. The findings indicate ongoing challenges in terms of subject composition, algorithm robustness and generalization, and algorithm feasibility for practical applications. Moreover, it suggests integrating the strengths of both MB and MF approaches to improve existing algorithms. Therefore, future research should further explore personalized MB-MF combination methods incorporating deep learning, attention mechanisms, muscle synergy features, motor unit features, and closed-loop feedback to achieve precise, real-time, and long-duration prediction of multi-joint complex movements, while further refining the transfer learning strategy for rapid algorithm deployment across days and subjects. Overall, this review summarizes the current research status, significant findings, and challenges, aiming to inspire future research on predicting upper limb motion intentions based on sEMG.

The motor imagery brain-computer interface (MI-BCI) based on electroencephalography (EEG) is a widely used human-machine interface paradigm. However, due to the non-stationarity and individual differences among subjects in EEG signals, the decoding accuracy is limited, affecting the application of the MI-BCI. In this paper, we propose the EISATC-Fusion model for MI EEG decoding, consisting of inception block, multi-head self-attention (MSA), temporal convolutional network (TCN), and layer fusion. Specifically, we design a DS Inception block to extract multi-scale frequency band information. And design a new cnnCosMSA module based on CNN and cos attention to solve the attention collapse and improve the interpretability of the model. The TCN module is improved by the depthwise separable convolution to reduces the parameters of the model. The layer fusion consists of feature fusion and decision fusion, fully utilizing the features output by the model and enhances the robustness of the model. We improve the two-stage training strategy for model training. Early stopping is used to prevent model overfitting, and the accuracy and loss of the validation set are used as indicators for early stopping. The proposed model achieves within-subject classification accuracies of 84.57% and 87.58% on BCI Competition IV Datasets 2a and 2b, respectively. And the model achieves cross-subject classification accuracies of 67.42% and 71.23% (by transfer learning) when training the model with two sessions and one session of Dataset 2a, respectively. The interpretability of the model is demonstrated through weight visualization method.

Steady-state visual evoked potential (SSVEP), one of the most popular electroencephalography (EEG)-based brain-computer interface (BCI) paradigms, can achieve high performance using calibration-based recognition algorithms. As calibration-based recognition algorithms are time-consuming to collect calibration data, the least-squares transformation (LST) has been used to reduce the calibration effort for SSVEP-based BCI. However, the transformation matrices constructed by current LST methods are not precise enough, resulting in large differences between the transformed data and the real data of the target subject. This ultimately leads to the constructed spatial filters and reference templates not being effective enough. To address these issues, this paper proposes multi-stimulus LST with online adaptation scheme (ms-LST-OA)., Methods: The proposed ms-LST-OA consists of two parts. Firstly, to improve the precision of the transformation matrices, we propose the multi-stimulus LST (ms-LST) using cross-stimulus learning scheme as the cross-subject data transformation method. The ms-LST uses the data from neighboring stimuli to construct a higher precision transformation matrix for each stimulus to reduce the differences between transformed data and real data. Secondly, to further optimize the constructed spatial filters and reference templates, we use an online adaptation scheme to learn more features of the EEG signals of the target subject through an iterative process trial-by-trial., Results: ms-LST-OA performance was measured for three datasets (Benchmark Dataset, BETA Dataset, and UCSD Dataset). Using few calibration data, the ITR of ms-LST-OA achieved 210.01±10.10 bits/min, 172.31±7.26 bits/min, and 139.04±14.90 bits/min for all three datasets, respectively., Conclusion: Using ms-LST-OA can reduce calibration effort for SSVEP-based BCIs.

Machine vision and artificial intelligence hold promise across healthcare applications. In this paper, we focus on the emerging research direction of infant action recognition, and we specifically consider the task of reaching which is an important developmental milestone. We develop E-babyNet, a lightweight yet effective neural-network-based framework for infant action recognition that leverages the spatial and temporal correlation of bounding boxes of infants' hands and objects to reach for to determine the onset and offset of the reaching action. E-babyNet consists of two main layers based on two LSTM and a Bidirectional LSTM (BiLSTM) model, respectively. The first layer provides a pre-evaluation of the reaching action for each hand by providing onset and offset keyframes. Then, the biLSTM model merges the previous outputs to deliver an outcome of the reaching actions detection for each frame including the reaching hand. We evaluated our approach against four other lightweight structures using a dataset comprising 5,865 annotated images resulting in 16,337 bounding boxes from 375 distinctive infant reaching actions performed while sitting by different subjects in unconstrained (home/clinic) environments. Results illustrate the effectiveness of our approach and ability to provide reliable reaching action detection and offer onset and offset keyframes with a precision of one frame. Moreover, the biLSTM layer can handle the transition between reaching actions and help reduce false detections.

As an essential cognitive function, attention has been widely studied and various indices based on EEG have been proposed for its convenience and easy availability for real-time attention monitoring. Although existing indices based on spectral power of empirical frequency bands are able to describe the attentional state in some way, the reliability still needs to be improved. This paper proposed a subject-specific attention index based on the weighted spectral power. Unlike traditional indices, the ranges of frequency bands are not empirical but obtained from subject-specific change patterns of spectral power of electroencephalograph (EEG) to overcome the great inter-subject variance. In addition, the contribution of each frequency component in the frequency band is considered different. Specifically, the ratio of power spectral density (PSD) function in attentional and inattentional state is utilized to calculate the weight to enhance the effectiveness of the proposed index. The proposed subject-specific attention index based on the weighted spectral power is evaluated on two open datasets including EEG data of a total of 44 subjects. The results of the proposed index are compared with 3 traditional attention indices using various statistical analysis methods including significance tests and distribution variance measurements. According to the experimental results, the proposed index can describe the attentional state more accurately. The proposed index respectively achieves accuracies of 86.21% and 70.00% at the 1% significance level in both the t-test and Wilcoxon rank-sum test for two datasets, which obtains improvements of 41.38% and 20.00% compared to the best result of the traditional indices. These results indicate that the proposed index provides an efficient way to measure attentional state.

A wide range of emerging biomedical applications and clinical interventions rely on the ability to deliver living cells via hollow, high-aspect-ratio microneedles. Recently, microneedle arrays (MNA) have gained increasing interest due to inherent benefits for drug delivery; however, studies exploring the potential to harness such advantages for cell delivery have been impeded due to the difficulties in manufacturing high-aspect-ratio MNAs suitable for delivering mammalian cells. To bypass these challenges, here we leverage and extend our previously reported hybrid additive manufacturing (or "three-dimensional (3D) printing) strategy- i.e ., the combined the "Vat Photopolymerization (VPP)" technique, "Liquid Crystal Display (LCD)" 3D printing with "Two-Photon Direct Laser Writing (DLW)"-to 3D print hollow MNAs that are suitable for cell delivery investigations. Specifically, we 3D printed four sets of 650 μ m-tall MNAs corresponding to needle-specific inner diameters (IDs) of 25 μ m, 50 μ m, 75 μ m, and 100 μ m, and then examined the effects of these MNAs on the post-delivery viability of both dendritic cells (DCs) and HEK293 cells. Experimental results revealed that the 25 μ m-ID case led to a statistically significant reduction in post-MNA-delivery cell viability for both cell types; however, MNAs with needle-specific IDs ≥ 50 μ m were statistically indistinguishable from one another as well as conventional 32G single needles, thereby providing an important benchmark for MNA-mediated cell delivery., Competing Interests: CONFLICT OF INTEREST K. Rand-Yadin is Founding Director of SeeTrue Technology, LLC., which has a potential interested in commercializing the presented MNAs. CMJ is an employee of the VA Maryland Health Care System. The views in this paper do not reflect the views of the state of Maryland or the US Government. CMJ has equity positions in Cartesian Therapeutics and Barinthus Biotherapeutics.

This paper reports a microfabricated triaxial capacitive force sensor. The sensor is fully encapsulated with inert and biocompatible glass (fused silica) material. The sensor comprises two glass plates, on which four capacitors are located. The sensor is intended for subdermal implantation in fingertips and palms and providing tactile sensing capabilities for patients with paralyzed hands. Additional electronic components, such as passives and IC chips, can also be integrated with the sensor in a hermetic glass package to achieve an implantable tactile sensing system. Through attachment to a human palm, the sensor has been shown to respond appropriately to typical hand actions, such as squeezing or picking up a bottle.

Data and privacy laws, such as the GDPR, require mobile apps that collect and process the personal data of their citizens to have a legally-compliant policy. Since these mobile apps are hosted on app distribution platforms such as Google Play Store and App Store, the app publishers also require the app developers who wish to submit a new app or make changes to an existing app to be transparent about their app privacy practices regarding handling sensitive user data that requires sensitive permissions such as calendar, camera, microphone. To verify compliance with privacy regulators and app distribution platforms, the app privacy policies and permissions are investigated for consistency. However, little has been done to investigate GDPR completeness checking within the Android permission ecosystem. In this paper, we investigate the design and runtime approaches towards completeness checking of sensitive (‘dangerous’) Android permission policy declarations against GDPR. In this paper, we investigate the design and runtime approaches towards completeness checking of dangerous Android permission policy declarations against GDPR. Leveraging the MPP-270 annotated corpus that describes permission declarations in application privacy policies, six natural language processing and language modelling algorithms are developed to measure permission completeness during runtime while a proof of concept Class Unified Modeling Language Diagram (UML) tool is developed to generate GDPR-compliant permission policy declarations using UML diagrams during design time. This paper makes a significant contribution to the identification of appropriate permission policy declaration methodologies that a developer can use to target particular GDPR laws, increasing GDPR compliance by 12% in cases during runtime using BERT word embedding, measuring GDPR compliance in permission policy sentences, and a UML-driven tool to generate compliant permission declarations.

Link prediction in complex networks is a critical process aimed at uncovering hidden or potential connections among nodes. This technique is widely utilized in areas such as knowledge graphs. Current Graph Neural Networks (GNNs) often focus exclusively on determining whether nodes are connected or assessing the strength of these links by leveraging node attributes. They typically use network structure and attributes to develop node representations through neighborhood aggregation. However, these methods often overlook the intrinsic importance of the links themselves. This paper thoroughly examines the significance of link value based on network structure and introduces an innovative approach for estimating this value, and proposes a method that incorporates link value into both the formulation and training of a link prediction graph attention network. This integration not only boosts the accuracy of link predictions but also provides a theoretical basis for understanding the prediction results. We conducted extensive experiments in link prediction employing widely recognized benchmark datasets. The findings reveal that our proposed framework for link prediction exhibits commendable performance and generalization capabilities, and overall performance improved by an average of 1.2%, thereby establishing it as an effective baseline model.

The limited size of existing datasets and signal variability have hindered EEG-based emotion recognition. In this paper, we present a solution that simultaneously addresses both problems. Generative Adversarial Networks (GANs) have recently shown notable data augmentation (DA) success. Therefore, we leverage a GAN-based DA technique to enhance the robustness of our proposed emotion recognition model by synthetically increasing the size of our datasets. Moreover, we employ contrastive learning to improve the quality of the learned representations from EEG signals and mitigate the adverse impact of inter-subject and intra-subject variability in signals corresponding to the same stimuli or emotions. We do so by maximizing the similarity in the representation of such EEG signals. We perform EEG-based emotion classification using a Graph Neural Network (GNN), which learns the relationship between the extracted EEG features. We compare the proposed model with several recent state-of-the-art emotion recognition models on the DEAP and MAHNOB datasets. The experimental results demonstrate that the proposed model outperforms previous models with a 64.84% and 66.40% emotion classification accuracy on the test set of the DEAP dataset and a 66.98% and 71.69% emotion classification accuracy on the test set of the MAHNOB-HCI dataset for the valence and arousal emotional dimensions, respectively. We perform an ablation study to demonstrate how contrastive learning, GAN, and GNN contribute to improving the proposed solution’s performance.

Cognitive agents make systems autonomous through the process of decision automation by mining an existing knowledge repository at run time. These processes can often be highly compute intensive, and would thus run slowly on the low-power computing platforms typically seen in autonomous systems. This paper examines how knowledge be represented in a Q-table and proposes a novel fast algorithm to mine that knowledge based on constraints. We evaluate this approach for the knowledge mining process of a specific agent: Cognitively Enhanced Complex Event Processing (CECEP). Within CECEP, knowledge is represented using Cognitive Domain Ontologies (CDO), and is mined using situational inputs and constraints. This is a novel approach to store information and is able to accommodate CDOs with millions of solutions. To show that the approach can run on low power hardware in real-time, this algorithm was executed on two low-power minicomputing platforms - Intel’s NUC and Asus’s Tinker Board. At present, no other optimized CDO solvers can generate solutions on these platforms. The algorithm generated the same amount of solutions as a GPU-enabled optimized path-based forward checking CDO solver, while consuming around 7.7 and 5.15 times less energy (Joules) on the NUC and Tinker Board respectively.