Spatio-temporal Network Research Articles

Multi-channel spatio-temporal graph attention contrastive network for brain disease diagnosis.

Dynamic brain networks (DBNs) can capture the intricate connections and temporal evolution among brain regions, becoming increasingly crucial in the diagnosis of neurological disorders. However, most existing researches tend to focus on isolated brain network sequence segmented by sliding windows, and they are difficult to effectively uncover the higher-order spatio-temporal topological pattern in DBNs. Meantime, it remains a challenge to utilize the structure connectivity prior in the DBNs analysis. To address these problems, we propose a multi-channel spatio-temporal graph attention contrastive network for DBNs analysis. Specifically, we first construct dynamic brain functional networks from fMRI data with sliding windows, and embed the structural connectivity derived from diffusion tensor imaging (DTI) to the dynamic functional connectivity graph representation to construct multi-modal brain network. Second, we develop a multi-channel spatial attention contrastive network to extract topological features from the brain network within each time window. This network incorporates an intra-window graph contrastive constraint to enhance the discriminative ability of the extracted features. Moreover, temporal dependencies across windows are captured by integrating feature embeddings through a self-attention mechanism, and the inter-window recurrent contrastive constraint is devised to extract higher-order spatio-temporal topological features. Finally, a multi-layer perceptron (MLP) is used to classify the brain networks. Experiments on epilepsy and ADNI datasets show that our method outperforms several state-of-the-art approaches in diagnosing performance, and it provides discriminative graph features for related brain diseases.

A spatiotemporal autoregressive neural network interpolation method for discrete environmental factors

The spatiotemporal interpolation model is necessary for generating continuous distributions for spatiotemporally discrete sampling points. However, there remain challenges in spatiotemporal interpolation due to the complex spatiotemporal effect and the imprecise kernel functions. Here, we proposed a spatiotemporal autoregressive neural network interpolation model (STARNN) that incorporates adaptive spatiotemporal distance quantification and supervised learning. The 10-fold cross-validation modelling on sea surface temperature and coastal nutrients demonstrated that the STARNN model performs better than baseline models and can well depict reasonable spatiotemporal distributions for environmental factors. By proposing two stacked neural networks, the STARNN model can accurately integrate spatial and temporal distances and avoids subjective selection of the kernel function. This study developed a novel interpolation model for processing discrete spatiotemporal points by following the data-driven paradigm, which can offer decision support for simulating the spread of sea temperature anomalies and optimizing the distribution of water quality measurement stations.

STCNet: Spatio-Temporal Cross Network with subject-aware contrastive learning for hand gesture recognition in surface EMG.

This paper introduces the Spatio-Temporal Cross Network (STCNet), a novel deep learning architecture tailored for robust hand gesture recognition in surface electromyography (sEMG) across multiple subjects. We address the challenges associated with the inter-subject variability and environmental factors such as electrode shift and muscle fatigue, which traditionally undermine the robustness of gesture recognition systems. STCNet integrates a convolutional-recurrent architecture with a spatio-temporal block that extracts features over segmented time intervals, enhancing both spatial and temporal analysis. Additionally, a rolling convolution technique designed to reflect the circular band structure of the sEMG measurement device is incorporated, thus capturing the inherent spatial relationships more effectively. We further propose a subject-aware contrastive learning framework that utilizes both subject and gesture label information to align the representation of vector space. Our comprehensive experimental evaluations demonstrate the superiority of STCNet under aggregated conditions, achieving state-of-the-art performance on benchmark datasets and effectively managing the variability among different subjects. The implemented code can be found at https://github.com/KNU-BrainAI/STCNet.

STIT-Net- A Wavelet based Convolutional Transformer Model for Motor Imagery EEG Signal Classification in the Sensorimotor Bands.

Motor Imagery (MI) electroencephalographic (EEG) signal classification is a pioneer research branch essential for mobility rehabilitation. This paper proposes an end-to-end hybrid deep network "Spatio Temporal Inception Transformer Network (STIT-Net)" model for MI classification. Discrete Wavelet Transform (DWT) is used to derive the alpha (8-13) Hz and beta (13-30) Hz EEG sub bands which are dominant during motor tasks to enhance the performance of the proposed work. STIT-Net employs spatial and temporal convolutions to capture spatial dependencies and temporal information and an inception block with three parallel convolutions extracts multi-level features. Then the transformer encoder with self-attention mechanism highlights the similar task. The proposed model improves the classification of the Physionet EEG motor imagery dataset with an average accuracy of 93.52% and 95.70% for binary class in the alpha and beta bands respectively, and 85.26% and 87.34% for three class, for four class 81.95% and 82.66% were obtained in the alpha and beta band respective EEG based motor signals which is better compared to the results available in the literature. The proposed methodology is further evaluated on other motor imagery datasets, both for subject-independent and cross-subject conditions, to assess the performance of the model.

Remaining useful life prediction of lithium-ion battery based on AConvST-LSTM-Net-TL

Abstract The accurate prediction of the Remaining Useful Life (RUL) of lithium-ion batteries can significantly enhance the safety and reliability of these batteries, thereby reducing operational risks. However, numerous existing methodologies operate under the assumption that both training and testing data adhere to the same distribution pattern, hindering the application of successful laboratory-based models to different target batteries. Hence, this study introduces a transfer learning model for predicting lithium-ion battery life, utilizing an attention mechanism-driven convolutional neural network (ACNN) in conjunction with a spatiotemporal Long Short-Term Memory network (ST-LSTM). Initially, the deep charging process data is leveraged for battery capacity estimation, where a Generalized Additive Model (GAM) is employed to delineate the nonlinear trend of battery capacity decay and characterize the capacity degradation. Following feature extraction, the Maximum Mean Discrepancy (MMD) value is computed to assess the distribution disparity between the two domains. Subsequently, the AConvST-LSTM-Net computes capacity estimations, loss functions for both source and target domains, merging these with the MMD value to formulate a comprehensive loss function. In conclusion, by employing transfer learning, the refined model is adapted to the target domain dataset, enabling the accurate prediction of the RUL of lithium-ion batteries. This approach is validated using the NASA lithium-ion battery dataset and the National Big Data Alliance for New Energy Vehicles (NDANEV), demonstrating the superior accuracy and robustness of the AConvST-LSTM-Net-TL model in comparisn to other prediction methods.

Prediction of Foundation Pit Dewatering Water Level Based on Neural Network and Improved Kriging Interpolation

Based on a foundation pit project in Shenyang, a large number of water level detection data were collected. The research focus of this paper is to use these data to build a depth neural network model to predict the water level of foundation pit dewatering. The Spatio-temporal convolution neural network has two components for capturing temporal correlation and spatial correlation of nodes, which have strong capturing ability of spatio-temporal correlation. The construction of spatio-temporal convolution neural network model can realize the time domain prediction of water level at all measuring points at the same time. The Kriging interpolation model optimized by Dupuit dewatering formula is superior in the spatial prediction of dewatering water level in the whole study area. Compared with OK method, IDW method and SP method, the prediction performance of improved Kriging interpolation is improved by 33.02%, 38.26% and 44.96% respectively Based on STGCN-Kriging model, the average error is 0.71 m and the error percentage is 2.55%. Combining with STGCN network and improved Kriging interpolation, the discrete prediction results in time domain are mapped to the spatial distribution results of water level buried depth in time domain, and the spatial distribution of foundation pit dewatering water level is predicted in time domain.

Spatio-Temporal Transformer with Kolmogorov–Arnold Network for Skeleton-Based Hand Gesture Recognition

Manually crafted features often suffer from being subjective, having an inadequate accuracy, or lacking in robustness in recognition. Meanwhile, existing deep learning methods often overlook the structural and dynamic characteristics of the human hand, failing to fully explore the contextual information of joints in both the spatial and temporal domains. To effectively capture dependencies between the hand joints that are not adjacent but may have potential connections, it is essential to learn long-term relationships. This study proposes a skeleton-based hand gesture recognition framework, the ST-KT, a spatio-temporal graph convolution network, and a transformer with the Kolmogorov–Arnold Network (KAN) model. It incorporates spatio-temporal graph convolution network (ST-GCN) modules and a spatio-temporal transformer module with KAN (KAN–Transformer). ST-GCN modules, which include a spatial graph convolution network (SGCN) and a temporal convolution network (TCN), extract primary features from skeleton sequences by leveraging the strength of graph convolutional networks in the spatio-temporal domain. A spatio-temporal position embedding method integrates node features, enriching representations by including node identities and temporal information. The transformer layer includes a spatial KAN–Transformer (S-KT) and a temporal KAN–Transformer (T-KT), which further extract joint features by learning edge weights and node embeddings, providing richer feature representations and the capability for nonlinear modeling. We evaluated the performance of our method on two challenging skeleton-based dynamic gesture datasets: our method achieved an accuracy of 97.5% on the SHREC’17 track dataset and 94.3% on the DHG-14/28 dataset. These results demonstrate that our proposed method, ST-KT, effectively captures dynamic skeleton changes and complex joint relationships.

Predicting task bottlenecks in digital manufacturing enterprises based on spatio-temporal graph convolutional networks

Predicting task bottlenecks in digital manufacturing enterprises based on spatio-temporal graph convolutional networks

Enhancing High-Speed Railway Timetable Resilience: A Two-Level Spatiotemporal Network Model Focused on Disturbance Absorption

Abstract Enhancing the resilience of train timetables can effectively improve their resistance and recovery capabilities against operational disturbances, thereby ensuring the stable operation of the railway system and the quality of passenger transportation services. This paper defined timetable resilience as three parts: disturbance absorptive capacity, resistance capacity, and post-disturbance recovery capacity. These capacities were quantitatively evaluated using the buffer time of trains, the number of train delays at departures and arrivals, and the duration of delay state. The evaluation metrics of the three resilience capacities and the total travel time of trains were used as the objective to establish a spatiotemporal network model. Utilizing actual operational data from the Beijing–Shanghai high-speed railway in China, the model was validated through a case study. Sensitivity analysis of the model's key parameters was conducted under two experimental scenarios: routine operational disturbances and speed restrictions on specific sections. Results showed that our model can effectively reduce the delay time and the number of delays in both the routine operational disturbance scenario and the 300 km/h speed restriction scenario with a frequent disturbance value of 2 min. Moreover, the model's performance with 3-min frequent disturbance outperformed that with 2-min frequent disturbance in speed restriction at 250 km/h and 200 km/h, yielding a 30% improvement.

Spatiotemporal network dynamics and structural correlates in the human cerebral cortex in vitro.

Elucidating human cerebral cortex function is essential for understanding the physiological basis of both healthy and pathological brain states. We obtained extracellular local field potential recordings from cortical slices of neocortical tissue from refractory epilepsy patients. Multi-electrode recordings were combined with histological information, providing a two-dimensional spatiotemporal characterization of human cortical dynamics in control conditions and following modulation of the excitation/inhibition balance. Slices expressed spontaneous rhythmic activity consistent with slow wave activity, comprising alternating active (Up) and silent (Down) states (Up-duration: 0.08±0.03s, Down-duration: 2.62±2.12s, frequency: 0.75±0.39Hz). Up states propagated from deep to superficial layers, with faster propagation speeds than in other species (vertical: 64.6mm/s; horizontal: 65.9mm/s). GABAA blockade progressively transformed the emergent activity into epileptiform discharges, marked by higher firing rates, faster network recruitment and propagation, and infraslow rhythmicity (0.01Hz). This dynamical characterization broadens our understanding of the mechanistic organization of the human cortical network at the micro- and mesoscale.

Spatiotemporal Interaction Based Dynamic Adversarial Adaptive Graph Neural Network for Air-Quality Prediction

Air quality issues have become a major environmental concern, with severe air pollution significantly reducing air quality and posing threats to human health. Accurate air quality prediction is crucial for preventing individuals from suffering the detrimental effects of severe air pollution. Recently, deep learning methods based on spatiotemporal graph neural networks (GNNs) have made considerable progress in modeling the temporal and spatial dependencies within air quality data by integrating GNNs with sequential models. Unfortunately, previous work often treats temporal and spatial dependencies as independent components, neglecting the intricate interactions between them. This oversight prevents the models from fully exploiting the complex spatiotemporal dependencies in the data, adversely affecting their predictive performance. To address these issues, we propose a general spatiotemporal interaction framework for air quality prediction. This framework models the bidirectional interactions between temporal and spatial dependencies in a data-driven manner. Furthermore, we designed a spatiotemporal feature extraction module and a dynamic adversarial adaptive graph learning module based on this framework. We introduce the Spatial-Temporal Interaction based Dynamic Adversarial Adaptive Graph Neural Network, capable of capturing the complex interactions between spatiotemporal dependencies and learning the dynamic spatial topology among sites by incorporating the competitive optimization concept of generative adversarial networks. Extensive experiments on two real-world datasets demonstrate the effectiveness of the proposed method, outperforming existing baseline models.

Needle tracking and segmentation in breast ultrasound imaging based on spatio-temporal memory network

Needle tracking and segmentation in breast ultrasound imaging based on spatio-temporal memory network

A traffic prediction method for missing data scenarios: graph convolutional recurrent ordinary differential equation network

Traffic prediction plays an increasingly important role in intelligent transportation systems and smart cities. Both travelers and urban managers rely on accurate traffic information to make decisions about route selection and traffic management. Due to various factors, both human and natural, traffic data often contains missing values. Addressing the impact of missing data on traffic flow prediction has become a widely discussed topic in the academic community and holds significant practical importance. Existing spatiotemporal graph models typically rely on complete data, and the presence of missing values can significantly degrade prediction performance and disrupt the construction of dynamic graph structures. To address this challenge, this paper proposes a neural network architecture designed specifically for missing data scenarios—graph convolutional recurrent ordinary differential equation network (GCRNODE). GCRNODE combines recurrent networks based on ordinary differential equation (ODE) with spatiotemporal memory graph convolutional networks, enabling accurate traffic prediction and effective modeling of dynamic graph structures even in the presence of missing data. GCRNODE uses ODE to model the evolution of traffic flow and updates the hidden states of the ODE through observed data. Additionally, GCRNODE employs a data-independent spatiotemporal memory graph convolutional network to capture the dynamic spatial dependencies in missing data scenarios. The experimental results on three real-world traffic datasets demonstrate that GCRNODE outperforms baseline models in prediction performance under various missing data rates and scenarios. This indicates that the proposed method has stronger adaptability and robustness in handling missing data and modeling spatiotemporal dependencies.

Multivariate time series weather forecasting model using integrated secondary decomposition and Self-Attentive spatio-temporal learning network

<p>Weather forecasting is an essential but complex task because of its substantial influence on human life and advanced atmospheric dynamics. In recent years, deep learning-based techniques have gain high attention for weather forecasting. Nevertheless, current forecasting models ignore the interdependent relationship between variables across regions and primarily examine temporal patterns of weather data. In this work, a new Multivariate time series weather forecasting model is proposed using integrated secondary decomposition and Self-Attentive spatio-temporal learning network (SASTLNet). Particularly, the weather data is filtered using a Singular Spectrum Decomposition with Fuzzy Entropy (SSD-FE) for removing the irrelevant components and screening the component containing vital information. Moreover, the spatiotemporal fluctuations in the meteorological variables are investigated using an enhanced empirical mode decomposition (E3MD) method and an SASTLNet model. The self-attention block of the SASTLNet model is discovered using ladder format to lower the processing costs while representing the temporal and spatial features through single memory cell. The simulation results prove that the suggested model can outperforms the baseline models in terms of efficacy and accuracy.</p>

Hyperbolic Sine Function Control-Based Finite-Time Bipartite Synchronization of Fractional-Order Spatiotemporal Networks and Its Application in Image Encryption

This work is devoted to the hyperbolic sine function (HSF) control-based finite-time bipartite synchronization of fractional-order spatiotemporal networks and its application in image encryption. Initially, the addressed networks adequately take into account the nature of anisotropic diffusion, i.e., the diffusion matrix can be not only non-diagonal but also non-square, without the conservative requirements in plenty of the existing literature. Next, an equation transformation and an inequality estimate for the anisotropic diffusion term are established, which are fundamental for analyzing the diffusion phenomenon in network dynamics. Subsequently, three control laws are devised to offer a detailed discussion for HSF control law’s outstanding performances, including the swifter convergence rate, the tighter bound of the settling time and the suppression of chattering. Following this, by a designed chaotic system with multi-scroll chaotic attractors tested with bifurcation diagrams, Poincaré map, and Turing pattern, several simulations are pvorided to attest the correctness of our developed findings. Finally, a formulated image encryption algorithm, which is evaluated through imperative security tests, reveals the effectiveness and superiority of the obtained results.

Traffic Flow Prediction for Highway Based on Multi-Task Spatiotemporal Graph Network

Traffic Flow Prediction for Highway Based on Multi-Task Spatiotemporal Graph Network

Season and Flow Drive Productivity of a Regulated River

Flow regimes of river ecosystems worldwide have undergone substantial changes because of water resource development, altering the way in which organic matter is generated and cycled throughout entire river catchments. Flow–ecology studies have focused on structural variables measured at small spatial scales. This creates a challenging mismatch when applying adaptive flow management for ecosystem functioning at a catchment or regional scale. Here, we sought to inform flow management by evaluating the drivers of ecosystem metabolism in a regulated river and assessing our ability to predict metabolism at unmonitored locations. We estimated rates of ecosystem metabolism from high-frequency monitoring of dissolved oxygen concentration at eight sites on the Lachlan River of Australia’s Murray–Darling Basin. We then applied a spatio-temporal stream network model to predict metabolism at unmonitored locations using only remotely sensed and gauging station predictor variables. Gross primary productivity (GPP) was higher at sites with lower mean annual discharge, and strong seasonal patterns in rates of productivity tended to be disrupted by rising flows. Similarly, ecosystem respiration (ER) was higher at sites with lower mean annual discharge and lower annual flow variation, but increased slightly in response to higher daily flows. Predictions at validation sites were generally accurate, albeit with substantial site-to-site variation. Our results suggest that flow changes may have altered metabolic rates from conditions prior to water abstraction and dam construction. These findings will assist in managing flows for ecosystem function outcomes and support extrapolation from monitored sites to the broad scales required for evaluating catchment-scale outcomes of river management.

Gradient-boosted spatiotemporal neural network for simulating underground hydrogen storage in aquifers

Underground hydrogen storage (UHS) in aquifers has emerged as a viable solution to address the seasonal mismatch between supply and demand in renewable energy. Numerical simulation of the UHS serves as a crucial foundation for optimizing storage operations and conducting system risk assessments. However, numerical simulation methods employed for these purposes often demand substantial data, making data collection challenging and computationally expensive, especially in scenarios involving the coupling of multiple physical fields. Deep learning serves as an effective tool in resolving this challenge. Here, we proposed a spatiotemporal neural network architecture with gradient enhancement, denoted as gradient-boosted spatiotemporal neural network (GSTNN) and its variant GSTNN-s. The GSTNN combines a convolutional neural network (CNN), a long short-term memory network (LSTM), and an autoencoder architecture. To incorporate physical constraints into the network, the spatiotemporal gradient operators from the gas-water seepage and gas convection-diffusion equations are introduced as regularization terms, imposing physics-informed constraints on the training process in both temporal and spatial dimensions. In predicting the multiphase flow of UHS in both homogeneous and heterogeneous formations, GSTNN outperforms CNN and CNN-LSTM in terms of the accuracy of pressure, saturation and H2 concentration fields. In terms of predicting UHS in formations with different permeabilities and porosities, GSTNN-s demonstrates improved performance as well. The proposed GSTNN architecture is promising in improving the efficiency of UHS numerical simulation, and has great potential to be applied for optimizing UHS operations in the future.

Spatiotemporal graph convolutional network using sparse monitoring data for accurate water-level reconstruction in urban drainage systems

Spatiotemporal graph convolutional network using sparse monitoring data for accurate water-level reconstruction in urban drainage systems

Multi-region infectious disease prediction modeling based on spatio-temporal graph neural network and the dynamic model.

Human mobility between different regions is a major factor in large-scale outbreaks of infectious diseases. Deep learning models incorporating infectious disease transmission dynamics for predicting the spread of multi-regional outbreaks due to human mobility have become a hot research topic. In this study, we incorporate the Graph Transformer Neural Network and graph learning mechanisms into a metapopulation SIR model to build a hybrid framework, Metapopulation Graph Transformer Neural Network (M-Graphormer), for high-dimensional parameter estimation and multi-regional epidemic prediction. The framework effectively solves the problem that existing models may lose some hidden spatial dependencies in the data when dealing with the dynamic graph structure of the network due to human mobility. We performed multi-wave infectious disease prediction in multiple regions based on real epidemic data. The results show that the framework is capable of performing high-dimensional parameter estimation and accurately predicting epidemic transmission dynamics in multiple regions even with low data quality. In addition, we retrospectively extrapolate the temporal evolution patterns of contact rate under different interventions implemented in different regions, reflecting the dynamics of intervention intensity and the need for flexibility in adjusting interventions in different regions. To provide early warning of infectious disease transmission, we retrospectively predicted the arrival time of infectious diseases using data from the early stages of outbreaks.