Alzheimer's disease (AD) is the most prevalent form of dementia, and its early diagnosis is essential for slowing disease progression. Recent studies on multimodal neuroimaging fusion using MRI and PET have achieved promising results by integrating multi-scale complementary features. However, most existing approaches primarily emphasize cross-modal complementarity while overlooking the diagnostic importance of modality-specific features. In addition, the inherent distributional differences between modalities often lead to biased and noisy representations, degrading classification performance. To address these challenges, we propose a Collaborative Attention and Consistent-Guided Fusion framework for MRI and PET based AD diagnosis. The proposed model introduces a learnable parameter representation (LPR) block to compensate for missing modality information, followed by a shared encoder and modality-independent encoders to preserve both shared and specific representations. Furthermore, a consistency-guided mechanism is employed to explicitly align the latent distributions across modalities. Experimental results on the ADNI dataset demonstrate that our method achieves superior diagnostic performance compared with existing fusion strategies.

Non-contact Ballistocardiography (BCG) and Electroencephalography (EEG) are widely used, yet while correlations between EEG and Electrocardiography (ECG), and BCG and ECG, are established, the feature-level correlation between BCG and EEG remains underexplored. To address this research gap, We collected 5 -minute supine EEG, ECG, and EMFi sensor signals from 10 subjects. BCG signals were extracted using Empirical Mode Decomposition (EMD) and EEG signals via Independent Component Analysis (ICA). Heart Rate Variability (HRV) features were derived from the BCG, and time-frequency features were extracted from the EEG for correlation analysis. The study revealed significant correlation patterns. Notably, BCG-derived HRV features (SDNN, RMSSD) showed strong positive correlations with EEG Hjorth Complexity and Kurtosis, alongside a strong negative correlation with EEG Hjorth Mobility ($\vert r\vert >0.9, p<0.001$). These findings demonstrate novel featurelevel correlations between BCG and EEG. The observed patterns are consistent with the subjects' physiological states and align with established principles of cardio-encephalic coherence.

Motor Imagery Brain-Computer Interface (MIBCI) is one of the most widely used BCI paradigms. However, due to large inter-individual variability in EEG signals, EEGbased pattern recognition models face significant challenges in cross-subject generalization. In this study, we posit that although cross-subject MI-BCI is fundamentally a cross-domain task, the population is not homogeneous; instead, latent subpopulations exist in which subjects share more consistent classification boundaries. To exploit this structure, we propose a Mixture-of-Experts (MoE) framework that automatically partitions training trials into latent groups and trains expert classifiers specialized for each group, while simultaneously learning a gating network that assigns test trials to the most suitable expert(s). This design enables the system to adapt to subpopulation structure, mitigate negative transfer from dissimilar subjects, and better model inter-subject heterogeneity. Evaluations on two public MI-EEG datasets (EEGMMIDB and OpenBMI) using k-fold cross-subject protocols demonstrate that our MoE approach significantly improves classification accuracy compared to most baseline models.

Alzheimer's Disease (AD) is the most prevalent neurodegenerative disorder in aging populations, posing a significant and escalating burden on global healthcare systems. While Multi-Tusk Learning (MTL) has emerged as a powerful computational paradigm for modeling longitudinal AD data, existing frameworks do not account for the time-varying nature of feature correlations. To address this limitation, we propose a novel MTL framework, named Feature Similarity Laplacian graph Multi-Task Learning (MTL-FSL). Our framework introduces a novel Feature Similarity Laplacian (FSL) penalty that explicitly models the time-varying relationships between features. By simultaneously considering temporal smoothness among tasks and the dynamic correlations among features, our model enhances both predictive accuracy and biological interpretability. To solve the non-smooth optimization problem arising from our proposed penalty terms, we adopt the Alternating Direction Method of Multipliers (ADMM) algorithm. Experiments conducted on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset demonstrate that our proposed MTLFSL framework achieves state-of-the-art performance, outperforming various baseline methods. The implementation source can be found at https://github.com/huatxxx/MTL-FSL.

Severity assessment of Parkinson's disease (PD) using wearable sensors offers an effective, objective basis for clinical management. However, general-purpose time series models often lack pathological specificity in feature extraction, making it difficult to capture subtle signals highly correlated with PD. Furthermore, the temporal sparsity of PD symptoms causes key diagnostic features to be easily “diluted” by traditional aggregation methods, further complicating assessment. To address these issues, we propose the Multi-scale Frequency-Aware Adversarial Multi-Instance Network (MFAM). This model enhances feature specificity through a frequency decomposition module guided by medical prior knowledge. Furthermore, by introducing an attention-based multi-instance learning (MIL) framework, the model can adaptively focus on the most diagnostically valuable sparse segments. We comprehensively validated MFAM on both the public PADS dataset for PD versus differential diagnosis (DD) binary classification and a private dataset for four-class severity assessment. Experimental results demonstrate that MFAM outperforms general-purpose time series models in handling complex clinical time series with specificity, providing a promising solution for automated assessment of PD severity.

Transformer-based deep neural networks have achieved remarkable success across various computer vision tasks, largely attributed to their long-range self-attention mechanism and scalability. However, most transformer architectures embed images into uniform, grid-based vision tokens, neglecting the underlying semantic meanings of image regions, resulting in suboptimal feature representations. To address this issue, we propose Fuzzy Token Clustering Transformer (FTCFormer), which incorporates a novel clustering-based downsampling module to dynamically generate vision tokens based on the semantic meanings instead of spatial positions. It allocates fewer tokens to less informative regions and more tokens to represent semantically important regions, regardless of their spatial adjacency or shape irregularity. To further enhance feature extraction and representation, we propose a Density Peak Clustering-Fuzzy K-Nearest Neighbor (DPC-FKNN) mechanism for clustering center determination, a Spatial Connectivity Score (SCS) for token assignment, and a channel-wise merging (Cmerge) strategy for token merging. Extensive experiments on 32 datasets across diverse domains validate the effectiveness of FTCFormer on image classification, showing consistent improvements over the TCFormer baseline, achieving gains of improving 1.43% on five fine-grained datasets, 1.09% on six natural image datasets, 0.97% on three medical datasets and 0.55% on four remote sensing datasets. The code is available at: https://github.com/BaoBao0926/FTCFormer/tree/main.

Parkinson’s disease (PD) is a progressive neurodegenerative disorder that significantly diminishes patients’ quality of life. Early and accurate diagnosis is critical for reducing both individual and societal burdens. Although current diagnostic methods can effectively differentiate between PD patients and healthy individuals, they tend to ignore the diversity of PD symptoms and the differences with other similar diseases, such as essential tremor or multiple system atrophy, leading to a higher risk of misdiagnosis. Additionally, existing supervised learning methods rely on subjective labeling by physicians, which is both time-consuming and subjective. To overcome these limitations, we collect multi-sensor activity data from 102 participants in free-living environments, and propose a novel self-supervised learning framework that redefines PD diagnosis as an anomaly detection problem. Additionally, we utilize two large public PD datasets as external cohorts to verify their validity. Extensive experiments demonstrate that our framework not only learns more discriminative features but also significantly enhances the model’s generalization ability, providing a promising solution to reduce misdiagnosis in PD.

This paper proposes a novel deep learning architecture, EEG-TBSANet, for automatic detection of epileptic seizures from electroencephalogram (EEG) signals. The model integrates temporal convolutional networks (TCN), bidirectional long short-term memory (BiLSTM) networks, and self-attention (SA) mechanisms in one single framework which allows the mechanism to locally and globally extract short and long temporal features while adaptively focusing on relevant components of the signals to detect seizures. EEG-TBSANet was fully tested on three publicly available benchmark datasets: Guinea-Bissau, Bonn, and CHB-MIT, which have different characteristics of signal types and acquisition scenarios. The model achieved classification accuracies of 98.02%, 98.57%, and 99.40% for the corresponding datasets under competitive conditions with several significant baseline and ablation models, including TCN-SA, LSTM-GRU, CNN-BiLSTM, TCN-BiLSTM, MDFLN. Additionally, an extensive ablation study confirmed the critical role of each architectural component in enhancing detection performance. The results show the model’s strong generalization ability across datasets with varied clinical and technical conditions. These results demonstrate that EEG-TBSANet offers a robust and generalizable solution for EEG-based epileptic seizure detection.

The prediction of fertilization rates is a critical area of research in the agricultural field and is essential for ensuring global food security. With the ongoing expansion of the global population and the escalating repercussions of climate change, precise crop fertilization rate predictions have become paramount. This is because accurate predictions can optimize resource allocation and improve agricultural productivity. Moreover, they can provide scientific support for policy-making and agricultural input management, thereby promoting sustainable agricultural development. Despite its importance, the complexity of agricultural systems, which is influenced by multiple factors including climate, geography, soil conditions, and management practices, poses significant challenges to prediction accuracy. In this paper, we propose a deep learning framework based on Graph Neural Networks (GNNs) that effectively incorporates geographical knowledge and multi-dimensional feature information. By modeling spatial relationships through graph structures (nodes and edges), our framework enhances fertilization rate prediction accuracy. We validate the model using two datasets of different scales. The results demonstrate excellent predictive performance across all datasets and strong scalability, highlighting its potential for agricultural fertilization rate prediction.

Early detection of breast cancer is essential for timely therapy. Ultrasound (US) helps differentiate benign from malignant breast masses, and the newest generation—Automated Breast Ultrasound (ABUS)—offers operator-independent acquisition, more accurate lesion-size estimation, and whole-breast coverage compared with Handheld Ultrasound (HHUS). Yet most computer-aided US studies still rely on HHUS because exploiting ABUS’s multi-view volumes is technically demanding. Existing ABUS models typically use convolutional neural networks (CNNs) and merge views by stacking them into 2.5-D inputs, a strategy that obscures how each view contributes to the final prediction.We introduce MVHT (Multi-view Hybrid Transformer), a transformer-based framework that explicitly fuses the multiple 2-D views extracted from an ABUS volume. To let the self-attention mechanism recognize the origin of each view, we design normal-vector position embedding (NVPE)—a simple but effective positional encoding keyed to the view’s surface normal. MVHT was trained and tested on a private cohort of 352 multi-view ABUS scans (1,056 images) and surpassed CNN baselines, achieving 91.86% accuracy in benign-versus-malignant classification. On a dataset derived from the public TDSC-ABUS Challenge, MVHT also led the benchmarks with an accuracy of 71.61% . Extensive ablation experiments confirm that NVPE boosts performance.Taken together, these results indicate that MVHT can enhance the diagnostic value of ABUS and has strong potential for clinical deployment in large-scale breast-cancer screening.

With the widespread deployment of sensors in portable devices such as smartphones and smartwatches, sensor-based human activity recognition (HAR) has emerged as a key research focus in the field of ubiquitous computing. However, the sequential nature of sensor data makes manual annotation extremely labor-intensive and time-consuming, severely limiting the scalability of traditional supervised learning approaches. To address this challenge, self-supervised learning (SSL) has become a highly promising alternative by enabling the extraction of effective feature representations from large amounts of unlabeled data.In this work, we propose a novel attention-guided time-channel masking strategy for self-supervised learning in HAR tasks. Unlike traditional random masking methods, our approach utilizes multi-head attention mechanisms to dynamically select and mask the most discriminative portions of the sequence, thereby guiding the model to learn deeper structural patterns and salient features within the data. Extensive experiments conducted on three public datasets — MotionSense, USC-HAD, and UCI-HAR — demonstrate that our method significantly outperforms both traditional random masking and fully supervised baselines in terms of macro-averaged F1-score and classification accuracy. Furthermore, ablation studies validate the robustness and effectiveness of the proposed attention-guided masking strategy under different masking ratios, with particularly notable advantages observed at lower masking rates. These results collectively confirm the effectiveness and potential applicability of our method for enhancing self-supervised representation learning in HAR tasks.

Agriculture is crucial for food production and rural economies, yet pest infestations significantly reduce crop yields. Traditional pest management heavily relies on manual monitoring and expert experience, limiting its automation potential. While recent advances in automated pest detection have improved early identification, integrating expert knowledge to support automated decision-making remains challenging. To address these challenges, this study proposes an intelligent diagnostic framework integrating object detection, retrieval-augmented technology, and Large Language Models (LLMs). In the object detection stage, a customized YOLOv8 model is optimized through data augmentation and an Adaptive Feature Pyramid Network, achieving pest identification and lightweight optimization. Subsequently, automated online retrieval technology extracts relevant pest control information, while a locally deployed DeepSeek LLM analyzes, filters, and summarizes the retrieved content to generate professional recommendations. Experimental results validate the effectiveness of the framework. In the information processing stage, retrieval-augmented LLMs effectively mitigate the "hallucination" phenomenon, significantly enhancing the professionalism and credibility of generated recommendations. Concurrently, the optimized object detection stage achieves remarkable results. On the Pest24 dataset, the improved YOLOv8 model demonstrates a significant performance boost, with [email protected] increasing by 13%, model size reducing by 30%, and an inference speed of 275 FPS. Compared to standalone LLM-based systems, our proposed intelligent diagnostic framework demonstrates enhanced decision-making reliability through multi-stage collaborative optimization, addressing both model hallucination and detection efficiency. This work provides a scalable solution for intelligent pest management.

Deep Learning methodologies have demonstrated exceptional efficacy in sentiment classification tasks. However, their extended inference times often impede practical deployment, particularly in resource-constrained environments. This paper addresses the challenge of reducing inference time by introducing a novel in-GPU caching approach, termed SentireCache, specifically designed for sentiment classification tasks. While traditional caching methods with the cosine similarity measurement have shown some reduction in inference time, they suffer from low hit rates and accuracy. To overcome this limitation, we incorporate a token filtering mechanism based on saliency into the caching system, along with simplified similarity calculation methods. The effectiveness of our proposed approach is theoretically analyzed. Moreover, extensive experimentation is conducted to compare SentireCache with other state-of-the-art caching methods. The results demonstrate a significant 37.7% reduction in inference time with an average performance degradation of 4.69%.

The emergence of the Omicron variant in Shanghai in 2022 has highlighted the need for effective diagnostic tools for COVID-19. Recent studies have indicated that cough sounds generate distinctive features capable of distinguishing infected individuals from healthy ones. This study proposes a hybrid approach that combines machine learning and pre-trained models, utilizing audio features such as Mel-frequency cepstral coefficients and waveforms as inputs to train both types of models for accurate identification of COVID-19 patients and healthy individuals. The study employs a dataset collected from hospitals in Shanghai, China, comprising 78 participants, including COVID-19 positive and negative individuals. The proposed method demonstrates superior performance in diagnosing COVID-19 compared to existing mainstream machine learning algorithms. Furthermore, decisive important audio features for the COVID-19 positive classifier are identified via SHAP values for feature importance. Overall, the proposed approach achieves excellent diagnostic accuracy for COVID-19, outperforming current mainstream machine learning methods. With its multiple strengths in performance, speed, and usability, this algorithm shows great promise in enabling large-scale screening and aiding the containment of future widespread infections.

As the population rapidly ages, Alzheimer’s disease (AD), the most common form of dementia, urgently requires the identification of reliable structural brain biomarkers and the development of effective therapeutic strategies. Multiple multi-task learning (MTL) paradigms have been developed to enhance model generalization by sharing information between tasks to predict AD progression and accurately identify MRI-associated biomarkers. Unlike previous MTL approaches that consider only a single kind of cognitive score to predict the complicated AD progression over time, we have developed an innovative MTL method to deal with various cognitive scores simultaneously, with each focusing on different aspects of patient cognition. To effectively capture the intricate associations among different cognitive scores at multiple time points, we first propose an Adaptive Multiple Cognitive Objective Temporal (AMCOT) task-relationship binding penalty mechanism. This mechanism adaptively reveals temporal correlations between various cognitive scores at different time points and uses these relationships to predict cumulative disease progression accurately. To select the most informative MRI features in AD progression, we consider integrating the sparse group Lasso into our model. Our algorithms are designed to handle large datasets efficiently. Empirical evaluation on the Alzheimer’s disease dataset shows that our approach significantly outperforms existing state-of-the-art algorithms in both overall and individual task performance. Additionally, we applied stability selection techniques to identify stable MRI biomarkers and analyzed their temporal patterns to gain insights into AD progression. The implementation source can be found at https://github.com/XuanhanFan/MTL-AMCOT-BB.

The automatic detection of epilepsy based on electroencephalography (EEG) has been proven effective under feature learning models. However, the practical implementation often encounters challenges from noise contamination presented in EEG signals. To address issues related to noise interference, we proposed a robust feature learning model for EEG-based epilepsy detection using Manhattan distance and L1 regularization. Specifically, we introduced Manhattan distance to construct the weight of the L1 regularization term in LASSO and obtained epilepsy-related information in the spectrum components of the original EEG signals and the differentiated signals through the LASSO-based feature selection. We verified the performance of our proposed model using the public EEG dataset. The model achieved the best performance, outperforming competing models. In addition, we tested the performance under noise interference by simulating EEG signal noise, indicating that our model is robust in EEG-based epilepsy detection.

Parkinson’s disease (PD) is a neurodegenerative disorder characterized by tremors as its most typical symptom. Wearable accelerometer sensors, along with corresponding machine learning algorithms, can effectively assist in the diagnosis of PD tremors. However, due to the variations in disease progression and symptoms caused by individual differences among PD patients, it is challenging for existing algorithms to eliminate label noise and accurately identify and extract disease-related features across diverse patient data. In this study, we propose a Learnable Discriminative Instance Pool (LDIP) algorithm based on multi-instance learning, which integrates the concept of learnable shapelets. This method transforms the traditional DIP algorithm into a learnable instance pool that can be adaptively adjusted according to discriminative criteria, thereby enhancing the separability between different classes after bag mapping. We evaluated the proposed method on two clinical datasets using three different machine learning classifiers, achieving a maximum 73% accuracy for 5-class classification. The experimental results demonstrate that our proposed method consistently outperforms current baselines across various settings.

Parkinson’s disease (PD) is a progressive neurode-generative disorder characterized by motor symptoms such as tremors, rigidity, and bradykinesia. Accurate and early diagnosis is crucial for effective management and treatment. Some quantitative studies have combined wearable technology with machine learning methods, demonstrating a high potential for practical application. However, these studies mostly use single-location, single-sensor data collected from PD patients in clinical settings, neglecting the diversity of PD symptoms and the real-world application scenarios in free-living environments. This paper proposes an auxiliary diagnosis framework for PD based on multi-location, multi-sensor fusion, and unsupervised domain adaptation. The multi-location, multi-sensor fusion can mitigate the asymmetry of Parkinson’s symptoms, while unsupervised domain adaptation helps transfer in-hospital data to free-living environments without the need for manual labeling of the free-living data. Additionally, this paper designs a multi-head attention mechanism that focuses the disease classifier on sensors with strong feature discrimination and good distribution alignment. This experiment relies on wearable sensor data from 60 PD patients and 12 healthy controls, achieving an impressive accuracy of 90.46%, a precision of 88.28%, a recall of 88.09%, and an F1-score of 88.14%.

Wearable sensors combined with machine learning provide an effective solution for assessing Parkinson’s Disease (PD) severity. However, time-series data from wearable sensors often lack window-level labels for PD severity, resulting in weak supervision, which introduces the challenge of label noise. Additionally, patient variability causes distributional discrepancies, further complicating the learning process. To address these issues, we propose Adaptive Domain-Adversarial Multi-Instance Learning (ADAMIL), which combines and refines Multiple-Instance Learning (MIL) with domain-adversarial techniques. We improve traditional MIL by incorporating self-attention mechanisms and learnable positional encoding, enabling ADAMIL to capture temporal dependencies more effectively, thus making it better suited for mitigating label noise in weakly supervised time-series data. Furthermore, ADAMIL refines domain-adversarial learning to autonomously align latent distributions, ensuring robust domain-invariant feature learning without relying on predefined labels. Experimental results show that ADAMIL achieves 85.29% accuracy and 80.57% F1-score in fine-grained PD severity classification, outperforming existing methods. Notably, this performance is achieved using only a single wrist-worn sensor, underscoring its potential for practical use in clinical and home settings. The code is available at https://github.com/xzxzy12345XZY/ADAMIL.

Alzheimer’s disease (AD) is the most common dementia in today’s aging society. Accurately predicting its progress remains a major challenge. Multi-task learning methods are widely used in AD research to help understand the progression of AD by predicting cognitive performance and identifying key imaging biomarkers. Previous work has selected representative feature subsets from magnetic resonance imaging (MRI) features. The design of these models is based on the assumption that correlations are consistent across different tasks. Specifically, the model only focuses on a single cognitive score in each prediction and ignores the correlation between different cognitive scores. However, clinicians often use a combination of assessment scores and other tests to more comprehensively assess cognitive status and make a diagnosis. Combining scores from multiple cognitive assessments helps improve accurate predictions of disease progression. Previous research models have primarily focused on predicting a single cognitive score longitudinally. In this paper, we propose a multi-target, multi-task learning method that comprehensively considers the correlation between different cognitive scores and the relationship between longitudinal tasks to simultaneously predict multiple cognitive scores to more comprehensively capture the disease characteristics of development, thereby effectively predicting disease progression. We also adopt a structure matrix to explicitly represent the correlation between tasks, further improving the accuracy and interpretability of the model. Results from extensive experiments using the Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset show that our method exhibits balanced multi-target performance when dealing with three cognitive scores. Compared to models focusing on a single cognitive target score, our method performs better in the early prediction of cognitive scores.

Container handling optimization in ports significantly influences logistics chain efficiency and cost control, vital for economic benefits. Traditional research prioritizes quay crane (QC) scheduling, while truck dynamic scheduling often ties directly to specific QCs, prolonging QC operation times and affecting port throughput and efficiency. To tackle this, the paper introduces a dynamic truck dispatching algorithm with two task allocation strategies. Experimental results reveal our algorithm reduces total QC makespan by 10.8% in general when compared to traditional methods and has increased effectiveness in large-scale problems.

The ballistocardiograph (BCG) is a non-contact technology that monitors the heart and provides detailed cardiovascular parameters. Despite its broad applicability for long-term home monitoring due to Covid-19, BCG signals face challenges from positional changes, body movements, and system noise, which impact detection algorithms. In this paper, we propose a method for detecting inter-beat intervals (IBI) based on signal fusion technology. We utilize a Dynamic Bayesian Network (DBN) to integrate five heartbeat localization features extracted from BCG signals. Additionally, Generative Adversarial Networks (GANs) are used to assess signal quality and select correlated channels, improving heart rate monitoring accuracy. Experimental results demonstrate an average coverage of 95.21% and a mean squared error of 0.05. These results outperform those of methods without channel selection and single-channel BCG, indicating the potential for improving IBI estimation in multichannel BCG signal sensor systems.

Bees play an important role in agricultural production. However, beekeeping relies on experienced beekeepers to take time and effort to maintain the bee colonies. To lower the threshold of beekeeping and improve efficiency, we proposed a sound-based bee colony state analysis model using compact Mel frequency cepstral coefficient (MFCC) patterns. Facing high-dimensional bee colony sound signals, we obtained MFCCs from the signals and constructed a set of filters to extract MFCC-based compact features called compact MFCC patterns. After extracting compact features, the feature set was given to the support vector machine classifier. We recorded the sounds of bee colonies under normal conditions and in the absence of the queen bee to verify the proposed model. While significantly compressing the dimensionality of MFCCs, the model still achieved an accuracy score of 99.30% in distinguishing the presence of the queen bee. The extracted compact MFCC patterns effectively and compactly characterize the information related to the bee colony states in the bee colony sound signals, giving the model an excellent ability to discriminate the state of the bee colony.

Depression is a kind of mental illness that is harmful to the development of society. Electroencephalography (EEG) is a promising tool in the area of auxiliary diagnosing diseases. In this paper, we develop a mutual information-based least absolute shrinkage and selection operator (MI-LASSO) model to learn representative features from the power spectral density (PSD) ratio extracted from data. Specifically, MI-LASSO adds an adaptive weight based on mutual information to LASSO, which can discriminate weights of different features. Following the feature selection accomplished by MI-LASSO, the feature set is input into the classifier. We design a stacking ensemble classifier composed of support vector machine (SVM), adaptive boosting (AdaBoost), random forest (RF), and K-nearest neighbor (KNN). Compared to independent classifiers, stacking has a stronger ability to recognize depression. The proposed framework is validated on the open datasets: MODMA and the dataset from Hospital Universiti Sains Malaysia (HUSM). The best classification accuracy on MODMA achieved 99.025%. The best classification accuracy on the second dataset achieved 99.06%. The results indicate that our framework outperforms other EEG-based methods in the identification of depression. We conducted several experiments whose results demonstrate our framework can effectively assist in the diagnosis of depression based on EEG.

Motor imagery (MI) is a mental process extensively used in the experimental paradigm for brain-computer interfaces (BCIs) across various basic science and clinical research studies. Despite its widespread use, accurately decoding intentions from MI poses significant challenges due to the complex nature of brain patterns and the limited sample sizes typically available for machine learning. This paper introduces a Spatiotemporal Graph Neural Network (ST-GCN) designed for MI classification. First, the spatial-temporal convolution layer is used to extract features from raw EEG data, where mixed depthwise convolution extracts temporal features, followed by spatial filtering convolution that decomposes the EEG signal. A graph convolution module employing the max relative aggregator is then utilized to explore the relationships between the spatially decomposed EEG components. In the final step, under the combined supervision of cross-entropy and our proposed channel selection loss, the ST-GCN achieves feature extraction that enhances interclass dispersion and intraclass compactness. We compare ST-GCN with several benchmark EEG decoding methods on two MI datasets: the BCI Competition III Dataset IVa and the BCI Competition IV Dataset 1. ST-GCN outperforms the deep learning benchmark methods by achieving an accuracy of 78.11% and 71.94%, respectively, in 10-fold cross-validation.

The Internet of Things (IoT) represents a complex, dynamic environment where edge devices continuously optimize their policies to address a continual stream of tasks. Previous studies have typically relied on a rehearsal buffer containing data from past tasks or a known task identity to mitigate catastrophic forgetting. Our research, Prompt Pool-based Continual Reinforcement Learning (PPCRL), aims to create a more efficient memory system by expanding a single prompt into a prompt pool, allowing agents to automatically select a set of relevant prompts without needing task identity knowledge. Similar to prompt-based learning techniques, our approach utilizes a small trainable prompt pool to guide pre-trained models through sequential task learning systematically. This allows us to optimize prompts for guiding model predictions and effectively manage both shared and task-specific knowledge while maintaining model generalization. We conducted experiments on two multi-agent benchmarks where traditional methods suffer from significant performance degradation. In contrast, PPCRL demonstrates the capability to outperform baselines and exhibits high generalization ability.

Abstract We propose a comprehensive soccer match video analysis pipeline tailored for broadcast footage, which encompasses three pivotal stages: soccer field localization, player tracking, and soccer ball detection. Firstly, we introduce sports camera calibration to seamlessly map soccer field images from match videos onto a standardized two‐dimensional soccer field template. This addresses the challenge of consistent analysis across video frames amid continuous camera angle changes. Secondly, given challenges such as occlusions, high‐speed movements, and dynamic camera perspectives, obtaining accurate position data for players and the soccer ball is non‐trivial. To mitigate this, we curate a large‐scale, high‐precision soccer ball detection dataset and devise a robust detection model, which achieved the of 80.9%. Additionally, we develop a high‐speed, efficient, and lightweight tracking model to ensure precise player tracking. Through the integration of these modules, our pipeline focuses on real‐time analysis of the current camera lens content during matches, facilitating rapid and accurate computation and analysis while offering intuitive visualizations.

Multimodal learning is widely used in automated early diagnosis of Alzheimer's disease. However, the current studies are based on an assumption that different modalities can provide more complementary information to help classify the samples from the public dataset Alzheimer's Disease Neuroimaging Initiative (ADNI). In addition, the combination of modalities and different tasks are external factors that affect the performance of multimodal learning. Above all, we summrise three main problems in the early diagnosis of Alzheimer's disease: (i) unimodal vs multimodal; (ii) different combinations of modalities; (iii) classification of different tasks. In this paper, to experimentally verify these three problems, a novel and reproducible multi-classification framework for Alzheimer's disease early automatic diagnosis is proposed to evaluate and verify the above issues. The multi-classification framework contains four layers, two types of feature representation methods, and two types of models to verify these three issues. At the same time, our framework is extensible, that is, it is compatible with new modalities generated by new technologies. Following that, a series of experiments based on the ADNI-1 dataset are conducted and some possible explanations for the early diagnosis of Alzheimer's disease are obtained through multimodal learning. Experimental results show that SNP has the highest accuracy rate of 57.09% in the early diagnosis of Alzheimer's disease. In the modality combination, the addition of Single Nucleotide Polymorphism modality improves the multi-modal machine learning performance by 3% to 7%. Furthermore, we analyse and discuss the most related Region of Interest and Single Nucleotide Polymorphism features of different modalities.

Identifying and utilising various biomarkers for tracking Alzheimer’s disease (AD) progression have received many recent attentions and enable helping clinicians make the prompt decisions. Traditional progression models focus on extracting morphological biomarkers in regions of interest (ROIs) from MRI/PET images, such as regional average cortical thickness and regional volume. They are effective but ignore the relationships between brain ROIs over time, which would lead to synergistic deterioration. For exploring the synergistic deteriorating relationship between these biomarkers, in this paper, we propose a novel spatio-temporal similarity measure based multi-task learning approach for effectively predicting AD progression and sensitively capturing the critical relationships between biomarkers. Specifically, we firstly define a temporal measure for estimating the magnitude and velocity of biomarker change over time, which indicate a changing trend(temporal). Converting this trend into the vector, we then compare this variability between biomarkers in a unified vector space(spatial). The experimental results show that compared with directly ROI based learning, our proposed method is more effective in predicting disease progression. Our method also enables performing longitudinal stability selection to identify the changing relationships between biomarkers, which play a key role in disease progression. We prove that the synergistic deteriorating biomarkers between cortical volumes or surface areas have a significant effect on the cognitive prediction.

Recently, there have been a wide spectrum of multitask learning (MTL) methods developed to model Alzheimer’s disease (AD) progression. Typical MTL studies related cognitive ability prediction focus on modeling AD progression using high-quality clinical data such as MRI and cognitive scores. These studies follow a unified regularised MTL framework to process each follow-up data from patients over time. Beginning at baseline, the framework regards cognitive ability at each followup as a task and organise task relationship through temporal smoothness in cognitive ability. There is little attention on how to design feasible experimental protocols and normalisation for reliably evaluating those regularised MTL models. In this paper, we present an empirical analysis for investigate above issues. Four typical structural regularization approaches are revisited. Four issues affecting evaluation process of regularised MTL models are evaluated by experiments: 1) evaluation indicators, 2) repeated experimental times, 3) training data size and 4) number of tasks in MTL. The results demonstrate that regularised MTL models are capable of predicting AD progression with effectiveness, in many challenging cases of curse of dimensionality, data insufficiency or single MRI data input. One important finding is that MTL can effectively reduce the over-fitting risk of model, even with limited sample size. We also discover that the temporal smoothness assumption instead limits the performance of later tasks. It encourages us to revisit the relationship between patients’ cognitive ability changes between 2 and 3 years when using MTL to model AD progression.

Internet of Things (IoT) enabled technology has rapidly and efficiently facilitate healthcare diagnose and treatment with low-cost and lightweight devices. Big data generated from IoT offers valuable and crucial information to guide decision-making, improve patient outcomes, and decrease healthcare costs, etc. The workshop is aiming to provide an opportunity for researchers and practitioners from both academia and industry to present the state-of-the-art research and applications in utilizing IoT and big data technology for healthcare by presenting efficient scientific and engineering solutions, addressing the needs and challenges for integration with new technologies, and providing visions for future research and development.

Internet of Things (IoT) Wearable technology plays a crucial role in assisting the diagnosis of Parkinson’s disease (PD), and an efficient model for auxiliary diagnosis of the severity of PD can help reduce the workload for doctors. However, due to the influence of data collection environments and annotators, noisy label data is inevitable, which may have a negative impact on modeling the severity of PD. To address the above challenges, on the one hand, we collected a large number of activity signal data of Parkinson’s patients in free-living environments, and on the other hand, we proposed an efficient PD stage assessment framework, which includes a noisy label processing method to alleviate the noisy label negative impact. Specifically, we collected signal data from 15 healthy controls and 68 PD patients through 12 activities, and then we proposed a framework for noisy label detection and correction. The experimental results on real PD data sets demonstrated that the proposed framework achieve 75.9% accuracy in PD stage assessment and significantly improve the classification performance of different types of basic classifiers, which is better than other noisy label detection algorithms and other PD stage assessment frameworks. Overall, in this work, we focus on modeling PD severity in free-living environments using a single wearable sensor and reducing the negative impact of noisy label data to better help PD patients manage the disease.

An effective auxiliary diagnostic model for the severity of Parkinson’s disease (PD) could help hospitals reduce their workload, particularly in nations or regions where medical resources are limited. However, a critical challenge persists that hampers the progress of such endeavors. Previous studies have employed label propagation techniques that assign uniform labels to all activity signal segments of a patient, neglecting the complex expression of PD symptoms, thereby introducing label noise. To confront this challenge, we have collected an extensive set of PD activity signals from a clinical setting and have proposed an efficient and robust framework for assessing PD severity. Specifically, we gathered wearable device data on 14 daily activities from 70 PD patients, based on the Unified Parkinson’s Disease Rating Scale Part III. Our data analysis indicates that many segments within the activities were incorrectly labeled, significantly impairing the classification performance of the model. We introduced a novel framework based on Multi-Instance Learning with a Re-weighted Discriminative Instance Mapping (RDIM) to model PD auxiliary diagnosis, aiming to eliminate the impact of label noise present in the data. The results demonstrate that our framework achieves an accuracy of 80.88% in classifying the severity of PD, effectively addressing the label noise caused by coarse-grained label propagation.

In order to control the first wave of COVID-19 pandemic in 2020, many models have shown effectiveness in predicting the spread of new coronary pneumonia and the different interventions. However, few models can collect large amounts of high-quality real-time data faster under the premise of protecting privacy, considering the impact of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variant and the mass vaccination program as a new intervention. Therefore, we developed a mobile intelligent application that can collect a large amount of real-time data while protecting privacy and conducted a feasibility study by defining a new COVID-19 mathematical model SEMCVRD. By simulating different intervention measures, the prediction model of the mobile intelligent application used in this article simulates the epidemic situation in the U.K. as an example. The findings are as below: the optimal intervention strategy is to suppress the intervention at$P=3$(intervention intensity: the average number of contacts per person per day) before the end of March 2021, then gradually release the intervention intensity at a rate of$P+2$, and finally release the intensity to$P=9$in June 2021. The COVID-19 pandemic will end at the end of June 2021, when the total number of deaths will reach 128772. This strategy will be able to balance the tradeoff between loss of life and economic loss. Compared with the official statistics released by the U.K. government on May 31, 2021, our model can accurately predict the relative error rate of the total number of cases is less than 6.9%, and the relative error rate of the total number of deaths is less than 1%. Furthermore, the model is also suitable for collecting data from countries/regions around the world.

Domestic waste classification was incorporated into legal provisions recently in China. However, relying on manpower to detect and classify domestic waste is highly inefficient. To that end, in this article, we propose a multimodel cascaded convolutional neural network (MCCNN) for domestic waste image detection and classification. MCCNN combined three subnetworks (DSSD, YOLOv4, and Faster-RCNN) to obtain the detections. Moreover, to suppress the false-positive predicts, we utilized a classification model cascaded with the detection part to judge whether the detection results are correct. To train and evaluate MCCNN, we designed a large-scale waste image dataset (LSWID), containing 30 000 domestic waste multilabeled images with 52 categories. To the best of our knowledge, the LSWID is the largest dataset on domestic waste images. Furthermore, a smart trash can is designed and applied to a Shanghai community, which helped to make waste recycling more efficient. Experimental results showed a state-of-the-art performance, with an average improvement of 10% in detection precision.

Electroencephalogram (EEG) is widely used to diagnose many neurological and psychiatric brain disorders. The correct interpretation of EEG data is critical to avoid misdiagnosis. However, the analysis of EEG data requires trained specialists and may vary from expert to expert. Meanwhile, it can be challenging and time-consuming to assess the EEG data since these signals may last several hours or days. Therefore, rapid and accurate classification of EEG data may be a key step towards interpreting EEG records. In this study, a novel deep learning model with an end-to-end structure is proposed to distinguish normal and abnormal EEG signals automatically. For this purpose, we investigate the possibility of combining the core ideas of inception and residual architectures into a hybrid model to improve classification performance. We evaluated the proposed method through extensive experiments on a real-world dataset, and it shows feasibility and effectiveness. Compared to previous studies on the same data, our method outperforms other existing EEG signal methods. Thus, the proposed method can aid clinicians to automatically detect brain activity.

With the aggravation of the population aging problem, Parkinson’s disease (PD) and other neurodegenerative diseases of the elderly are not only a medical problem but also an important social problem. Therefore, early detection of PD is particularly important for reducing complications. Currently, the diagnosis of PD is assessed by specialized physicians through the Uniform PD Rating Scale (UPDRS). This limits the detection rate of PD and the timely assessment of disease progression to a certain extent. Moreover, with the development of artificial intelligence, machine learning has been widely and effectively applied to the assessment and monitoring of PD. Therefore, we use machine learning to distinguish between healthy people and PD patients based on UPDRS. In this paper, we collaborated with the First People’s Hospital of Yunnan Province to collect exercise data from 15 healthy individuals and 15 PD patients using wearable motion sensors. The analysis found that not all activities collected according to the UPDRS were useful. According to our proposed Indicators for distinguishing activities (I-DA) method as defined in this article, the most differentiated activities are found. Retain the activities that contain the most discriminative information, and use these activities to distinguish between healthy people and PD patients. We verify the effectiveness of this method through experiments. We use k-Nearest Neighbor (KNN), eXtreme Gradient Boosting (XGB), and Support Vector Machine (SVM) to execute the classification method. When the selected activities were taken as the whole data set rather than all activities according to our proposed Indicators for distinguishing activities (I-DA) method, the classification accuracy of KNN and XGB were improved by 5.10% and 2.4% respectively. The classification accuracy of SVM was improved by 12.07%. The experimental results show that the accuracy is significantly improved.

Alzheimer’s Disease (AD) is the most common dementia characterized by loss of brain function. Multi-tasking learning methods have been widely used to predict cognitive performance and select important imaging biomarkers in AD research. The temporal smoothness assumption, prevalent for modeling AD progression, means the difference between cognitive scores at two consecutive time points is relatively small. However, it’s not appropriate due to the presence of sample disturbance and the effectiveness of drug therapy. In addition, many multi-task learning methods select discriminative feature subset from MRI features, assuming that correlations between tasks are consistent, which ignores the complex intrinsic correlation structure of tasks. In this paper, we present a multi-task learning framework which utilizes generalized fused Lasso and generalized group Lasso (GFGGL for abbreviation) to model the disease progression with the complex intrinsic nonlinear structures of disease. The proposed framework is more flexible to utilize the inherent nonlinear relation of AD than existing methods for the reason of we represent the intrinsic structure as three correlation matrices which are functions of super parameters. The framework involves (1) two nonlinear structures of disease progression and (2) one nonlinear structure among tasks. An efficient optimization method is designed for the difficult optimization problem due to the presence of three nonsmooth penalties. Extensive experimental results using dataset from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) demonstrate the effectiveness of the proposed method.

Recent emergence of “microblogging” services has been driving cyber-physical social system (CPSS) as a hot topic in real-world applications. How to efficiently detect and recognise spam and fake accounts becomes an important task where it requires analysis of microblog user behavior and prediction of their activity. This article attempts to investigate this challenge by proposing a new strategy to effectively model microblogging user activity and dynamically predicting their activities for the CPSS applications. We first analysis and define a set of benchmarks for measuring microblogging user activeness in considering serval key dynamic attributes including change rate of microblogging numbers, user attentions, etc. Then, we build up a new dynamic microblogging user activity prediction model (DUAPM) based on three important characteristics: personal information, social relationship, and user interaction. Finally, an improved logical regression algorithm is proposed for training the model and predicting user activity. Under the evaluation of a sample dataset containing Sina Weibo 3621 users over 20 weeks, it shows that our model deliver average up to 3% higher prediction accuracy than other social media user activity prediction models using traditional logical regression and random forest algorithms. We also take out a CPSS case study of evaluating DUAPM models for analysis and prediction of Twitter users' activity over 16 countries. The results show that our model effectively reflects the distribution and trends of Twitter users' activity with different background and cultures.

Modeling and predicting progression of chronic diseases like Alzheimer's disease (AD) has recently received much attention. Traditional approaches in this field mostly rely on harnessing statistical methods into processing medical data like genes, MRI images, demographics, etc. Latest advances of machine learning techniques grant another chance of training disease progression models for AD. This trend leads on exploring and designing new machine learning techniques towards multi-modality medical and health dataset for predicting occurrences and modeling progression of AD. This paper aims at giving a systemic survey on summarizing and comparing several mainstream techniques for AD progression modeling, and discuss the potential and limitations of these techniques in practical applications. We summarize three key techniques for modeling AD progression: multi-task model, time series model and deep learning. In particular, we discuss the basic structural elements of most representative multi-task learning algorithms, and analyze a multi-task disease prediction model based on longitudinal time. Lastly, some potential future research direction is given.

Due to the many beneficial effects on physical and mental health and strong association with many fitness and rehabilitation programs, physical activity (PA) recognition has been considered as a key paradigm for Internet of Things healthcare. Traditional PA recognition techniques focus on repeated aerobic exercises or stationary PA. As a crucial indicator in human health, it covers a range of bodily movement from aerobics to anaerobic that may all bring health benefits. However, existing PA recognition approaches are mostly designed for specific scenarios and often lack extensibility for application in other areas, thereby limiting their usefulness. In this paper, we attempt to detect more gym PAs (GPAs) in addition to traditional PA using acceleration, A two layer recognition framework is proposed that can classify aerobic, sedentary, and free weight activities, count repetitions and sets for the free weight exercises, and in the meantime, measure quantities of repetitions and sets for free weight activities. In the first layer, a one-class support vector machine is applied to coarsely classify free weight and nonfree weight activities. In the second layer, a neural network is utilized for aerobic and sedentary activities recognition; a hidden Markov model is to provide a further classification in free weight activities. The performance of the framework was tested on ten healthy subjects (age: 30 ± 5; BMI: 25 ± 5.5 kg/m2; and body fat: 20.5 ± 5.4), and compared with some typical classifiers. The results indicate the proposed framework has better performance in recognizing and measuring GPAs than other approaches. The potential of this framework can be extended in supporting more types of PA recognition in complex applications.

As the rapid growth of social media technologies continues, Cyber-Physical-Social System (CPSS) has been a hot topic in many industrial applications. The use of “microblogging” services, such as Twitter, has rapidly become an influential way to share information. While recent studies have revealed that understanding and modelling microblog user behaviour with massive users’ data in social media are keen to success of many practical applications in CPSS, a key challenge in literatures is that diversity of geography and cultures in social media technologies strongly affect user behaviour and activity. The motivation of this article is to understand differences and similarities between microblogging users from different countries using social media technologies, and to attempt to design a Country-Level Micro-Blog User (CLMB) behaviour and activity model for supporting CPSS applications. We proposed a CLMB model for analysing microblogging user behaviour and their activity across different countries in the CPSS applications. The model has considered three important characteristics of user behaviour in microblogging data, including content of microblogging messages, user emotion index, and user relationship network. We evaluated CLBM model under the collected microblog dataset from 16 countries with the largest number of representative and active users in the world. Experimental results show that (1) for some countries with small population and strong cohesiveness, users pay more attention to social functionalities of microblogging service; (2) for some countries containing mostly large loose social groups, users use microblogging services as a news dissemination platform; (3) users in countries whose social network structure exhibits reciprocity rather than hierarchy will use more linguistic elements to express happiness in microblogging services.

Social Internet-of-Things (SIoT) environment comprises not only smart devices but also the humans who interact with these IoT devices. The benefits of such system are overshadowed due to the cyber security issues. A novel approach is required to understand the security implication under such a dynamic environment while taking both the social and technical aspects into consideration. This paper addressed such challenges and proposed a 3-D security modeling platform that can capture and model the security requirements in the SIoT environment. The modeling process is graphical notation based and works as a security extension to the Business Process Model and Notation. Still, it utilizes the latest 3-D game technology; thus, the security extensions are generated through the third dimension. Consequently, the introduction of security extensions will not increase the complexity of the original SIoT scenario, while keeping all the key information on the same platform. Together with the proposed security ontology, these comprehensive security notations created a unique platform that aims at addressing the ever complicated security issues in the SIoT environment.

Due to storage conditions and material's non‐planar shape, geometric distortion of the two‐dimensional content is widely present in scanned document images. Effective geometric restoration of these distorted document images considerably increases character recognition rate in large‐scale digitisation. For large‐scale digitisation of historical books, geometric restoration solutions expect to be accurate, generic, robust, unsupervised and reversible. However, most methods in the literature concentrate on improving restoration accuracy for specific distortion effect, but not their applicability in large‐scale digitisation. This study proposes an effective mesh based geometric restoration system (GRLSD) for large‐scale distorted historical document digitisation. In this system, an automatic mesh generation based dewarping tool is proposed to geometrically model and correct arbitrary warping historical documents. An XML‐based mesh recorder is proposed to record the mesh of distortion information for reversible use. A graphic user interface (GUI) toolkit is designed to visually display and manually manipulate the mesh for improving geometric restoration accuracy. Experimental results show that the proposed automatic dewarping approach efficiently corrects arbitrarily warped historical documents, with an improved performance over several state‐of‐the‐art geometric restoration methods. By using XML mesh recorder and GUI toolkit, the GRLSD system greatly aids users to flexibly monitor and correct ambiguous points of mesh for the prevention of damaging historical document images without distortions in large‐scale digitalisation.