Objective. Recent technological advances show the feasibility of fusing surface electromyography (sEMG) signals and movement data to predict lower limb ambulation intentions. However, since the invasive fusion of different signals is a major impediment to improving predictive performance, searching for a non-invasive (NI) fusion mechanism for lower limb ambulation pattern recognition based on different modal features is crucial. Approach . We propose an end-to-end sequence prediction model with NI dual attention temporal convolutional networks (NIDA-TCNs) as a core to elegantly address the essential deficiencies of traditional decision models with heterogeneous signal fusion. Notably, the NIDA-TCN is a weighted fusion of sEMG and inertial measurement units with time-dependent effective hidden information in the temporal and channel dimensions using TCN and self-attentive mechanisms. The new model can better discriminate between walking, jumping, downstairs, and upstairs four lower limb activities of daily living. Main results . The results of this study show that the NIDA-TCN models produce predictions that significantly outperform both frame-wise and TCN models in terms of accuracy, sensitivity, precision, F1 score, and stability. Particularly, the NIDA-TCN with sequence decision fusion (NIDA-TCN-SDF) models, have maximum accuracy and stability increments of 3.37% and 4.95% relative to the frame-wise model, respectively, without manual feature-encoding and complex model parameters. Significance . It is concluded that the results demonstrate the validity and feasibility of the NIDA-TCN-SDF models to ensure the prediction of daily lower limb ambulation activities, paving the way to the development of fused heterogeneous signal decoding with better prediction performance.

The early detection of Alzheimer’s disease (AD) in cognitively healthy individuals remains a major preclinical challenge. EEG is a promising tool that has shown effectiveness in detecting AD risk. Task-related EEG has been rarely used in Alzheimer’s disease research, as most studies have focused on resting-state EEG. An interpretable deep learning framework—Interpretable Convolutional Neural Network (InterpretableCNN)—was utilized to identify AD-related EEG features. EEG data were recorded during three cognitive task conditions, and samples were labeled based on APOE genotype and polygenic risk scores. A 100-fold leave-p%-subjects-out cross-validation (LPSO-CV) was used to evaluate model performance and generalizability. The model achieved an ROC AUC of 60.84% across the tasks and subjects, with a Kappa value of 0.22, indicating fair agreement. Interpretation revealed a consistent focus on theta and alpha activity in the parietal and temporal regions—areas commonly associated with AD pathology. Task-related EEG combined with interpretable deep learning can reveal early AD risk signatures in healthy individuals. InterpretableCNN enhances transparency in feature identification, offering a valuable tool for preclinical screening.

The early prediction of Alzheimer’s disease (AD) risk in healthy individuals remains a significant challenge. This study investigates the feasibility of task-state EEG signals for improving detection accuracy. Electroencephalogram (EEG) data were collected from the Multi-Source Interference Task (MSIT) and Sternberg Memory Task (STMT). Time–frequency features were extracted using the Multitaper method, followed by multidimensional reduction techniques. Subspace features (F24 and F216) were selected via t-tests and False Discovery Rate (FDR) multiple comparisons correction, and subsequently analyzed in the Time–Frequency Area Average Test (TFAAT) and Prefrontal Beta Time Series Test (PBTST). The experimental results reveal that the MSIT task achieves optimal cross-subject classification performance using the Support Vector Machine (SVM) approach with the TFAAT feature set, yielding a Receiver Operating Characteristic Area Under the Curve (ROC AUC) of 58%. Similarly, the Sternberg Memory Task demonstrates classification ability with the logistic regression model applied to the PBTST feature set, emphasizing the beta band power spectrum in the prefrontal cortex as a potential marker of AD risk. These findings confirm that task-state EEG provides stronger classification potential compared to resting-state EEG, offering valuable insights for advancing early AD prediction research.

Physiological indicators are increasingly employed in sports training. However, studies on surface electromyography (sEMG) primarily focus on the analysis of isometric contraction. Research on sEMG related to isotonic contraction, which is more relevant to athletic performance, remains relatively limited. This paper examines the changes in the isotonic contraction performance of the male upper arm muscles resulting from long-term basketball training using the sEMG metrics. We recruited basketball physical education (B-PE) and non-PE majors to conduct a controlled isotonic contraction experiment to collect and analyze sEMG signals. The sample entropy event detection method was utilized to extract the epochs of active segments of data. Subsequently, statistical analysis methods were applied to extract the key sEMG time domain (TD) and frequency domain (FD) features of isotonic contraction that can differentiate between professional and amateur athletes. Machine learning methods were employed to perform ten-fold cross-validation and repeated experiments to verify the effectiveness of the features across the different groups. This paper investigates the key features and channels of interest for categorizing male participants from non-PE and B-PE backgrounds. The experimental results show that the F12B feature group consistently achieved an accuracy of between 80% and 90% with the SVM2 model, balancing both accuracy and efficiency, which can serve as evaluation indices for isotonic contraction performance of upper limb muscles during basketball training. This has practical significance for monitoring isotonic sEMG features in sports and training, as well as for providing individualized training regimens.

The emergence of modern prosthetics controlled by bio-signals has been facilitated by AI and microchip technology innovations. AI algorithms are trained using sEMG produced by muscles during contractions. The data acquisition procedure may result in discomfort and fatigue, particularly for amputees. Furthermore, prosthetic companies restrict sEMG signal exchange, limiting data-driven research and reproducibility. GANs present a viable solution to the aforementioned concerns. GANs can generate high-quality sEMG, which can be utilised for data augmentation, decrease the training time required by prosthetic users, enhance classification accuracy and ensure research reproducibility. This research proposes the utilisation of a one-dimensional deep convolutional GAN (1DDCGAN) to generate the sEMG of hand gestures. This approach involves the incorporation of dynamic time wrapping, fast Fourier transform and wavelets as discriminator inputs. Two datasets were utilised to validate the methodology, where five windows and increments were utilised to extract features to evaluate the synthesised sEMG quality. In addition to the traditional classification and augmentation metrics, two novel metrics—the Mantel test and the classifier two-sample test—were used for evaluation. The 1DDCGAN preserved the inter-feature correlations and generated high-quality signals, which resembled the original data. Additionally, the classification accuracy improved by an average of 1.21–5%.

Wide-range application scenarios, such as industrial, medical, rescue, etc., are in various demand for human spatial positioning technology. However, the existing MEMS-based sensor positioning methods have many problems, such as large accuracy errors, poor real-time performance and a single scene. We focused on improving the accuracy of IMU-based both feet localization and path tracing, and analyzed three traditional methods. In this paper, a planar spatial human positioning method based on high-resolution pressure insoles and IMU sensors was improved, and a real-time position compensation method for walking modes was proposed. To validate the improved method, we added two high-resolution pressure insoles to our self-developed motion capture system with a wireless sensor network (WSN) system consisting of 12 IMUs. By multi-sensor data fusion, we implemented dynamic recognition and automatic matching of compensation values for five walking modes, with real-time spatial-position calculation of the touchdown foot, enhancing the 3D accuracy of its practical positioning. Finally, we compared the proposed algorithm with three old methods by statistical analysis of multiple sets of experimental data. The experimental results show that this method has higher positioning accuracy in real-time indoor positioning and path-tracking tasks. The methodology can have more extensive and effective applications in the future.