Browse Articles
Discover research articles across all indexed journals
Systematic data management for effective AI-driven decision support systems in robotic rehabilitation
Abstract Robotic rehabilitation is becoming a standard in post-stroke physical rehabilitation, and these setups, often coupled with virtual exercises, collect a large and finely grained amount of data about patients’ motor performance, in terms of kinematics and force interactions. Given the high resolution of data throughout the rehabilitation treatment, invaluable information is concealed, especially if oriented towards predictive systems and decision support systems. Nevertheless, a comprehensive understanding of how manipulating these datasets with machine-learning to produce such outputs is still missing. This study leverages comprehensive robotic-assisted rehabilitation data to systematically investigate clinical outcome predictions (FMA, ARAT and MI) and robot parameters suggestions based solely on kinematic and demographic data. Our method significantly outperforms conventional approaches on both tasks demonstrating the potential of systematic data handling in advancing rehabilitation practices. Moreover, under the explainable-AI policies, a focus on prediction power of variables and a clinical knowledge base of predicted outcome are provided.
Machine learning approaches for predicting the link of the global trade network of liquefied natural gas
With the rising geopolitical tensions, predicting future trade partners has become a critical topic for the global community. Liquefied natural gas (LNG), recognized as the cleanest burning hydrocarbon, plays a significant role in the transition to a cleaner energy future. As international trade in LNG becomes increasingly volatile, it is essential to assist governments in identifying potential trade partners and analyzing the trade network. Traditionally, forecasts of future mineral and energy resource trade networks have relied on similarity indicators (e.g., CN, AA). This study employs complex network theory to illustrate the characteristics of nodes and edges, as well as the evolution of global LNG trade networks from 2001 to 2020. Utilizing node and edge data from these networks, this research applies machine learning algorithms to predict future links based on local and global similarity-based indices (e.g., CN, JA, PA). The findings indicate that random forest and decision tree algorithms, when used with local similarity-based indices, demonstrate strong predictive performance. The reliability of these algorithms is validated through the Receiver Operating Characteristic Curve (ROC). Additionally, a graph attention network model is developed to predict potential links using edge and motif data. The results indicate robust predictive performance. This study demonstrates that machine learning algorithms—specifically random forest and decision tree—outperform in predicting links within the global LNG trade network based on local information proximity, while the graph attention network, a deep learning model, exhibits stable optimization and effective feature learning. These findings suggest that machine learning approaches hold significant promise for mineral trade network analysis.
Prototype ultrasonic desalination instrument and its performance test
AC-YOLO: A lightweight ship detection model for SAR images based on YOLO11
Synthetic Aperture Radar (SAR), renowned for its all-weather monitoring capability and high-resolution imaging characteristics, plays a pivotal role in ocean resource exploration, environmental surveillance, and maritime security. It has become a fundamental technological support in marine science research and maritime management. However, existing SAR ship detection algorithms encounter two major challenges: limited detection accuracy and high computational cost, primarily due to the wide range of target scales, indistinct contour features, and complex background interference. To address these challenges, this paper proposes AC-YOLO, a novel lightweight SAR ship detection model based on YOLO11. Specifically, we design a lightweight cross-scale feature fusion module that adaptively fuses multi-scale feature information, enhancing small target detection while reducing model complexity. Additionally, we construct a hybrid attention enhancement module, integrating convolutional operations with a self-attention mechanism to improve feature discrimination without compromising computational efficiency. Furthermore, we propose an optimized bounding box regression loss function, the Minimum Point Distance Intersection over the Union (MPDIoU), which establishes multi-dimensional geometric metrics to accurately characterize discrepancies in overlap area, center distance, and scale variation between predicted and ground truth boxes. Experimental results demonstrate that, compared with the baseline YOLO11 model, AC-YOLO reduces parameter count by 30.0% and computational load by 15.6% on the SSDD dataset, with an average precision (AP) improvement of 1.2%; on the HRSID dataset, the AP increases by 1.5%. This model effectively reconciles the trade-off between complexity and detection accuracy, providing a feasible solution for deployment on edge computing platforms. The source code for the AC-YOLO model is available at: https://github.com/He-ship-sar/ACYOLO.
Sleep and modern life: a population-based study
Study on the propagation law of hydraulic fractures in heterogeneous permeability reservoirs
The development of unconventional oil and gas resources is increasingly shifting toward heterogeneous reservoirs with complex permeability distributions, making the effective control of hydraulic fracture propagation patterns critical for optimizing production. To this end, this study establishes a 3D multilayered heterogeneous reservoir model using the finite element method to analyze fracture mechanisms. The impacts of permeability heterogeneous, injection rate, and fracturing fluid viscosity on fracture morphology are systematically investigated, and the elasticity coefficient method was used to evaluate the influence weights of each parameter.The main conclusions are as follows: (1) Permeability distribution is the core factor controlling the fracture propagation direction, with HPL dominating the extension path while MPL and LPL show limited efficiency. (2) An increase in the number of permeability layers inhibits the overall expansion of cracks, and the shape of the cracks gradually changes to rectangular. (3) Higher injection rates significantly expand fracture area, whereas fracturing fluid viscosity ≥50 mPa·s stabilizes fracture morphology. (4) The elastic coefficient method identifies injection rate, permeability heterogeneous, and fracturing fluid viscosity as the key control parameters in order. This work provides theoretical guidance for optimizing hydraulic fracturing parameters in complex geological settings.
Optimal intervention design for tonsillitis transmission via compartmental modeling with stability analysis and control strategies
Evaluation capacity building in a rural Victorian community service organisation: A formative evaluation
Community service organisations are increasingly required to report on outcomes and evaluate program delivery. While commonplace in clinical health settings, such work is novel to the community sector and can be challenging to undertake given resourcing and evaluation capacity constraints. These constraints are exacerbated for rural community service organisations that face additional resource pressures and lack organisational systems to support such work. This formative evaluation reports on an evaluation capacity building program within a rural place-based community service organisation in southwest Victoria, Australia. As part of the program, monitoring, evaluation, and learning pilots were implemented to support selected teams (N = 4 teams, N = 12 individuals) to increase confidence in outcomes measurement and reporting. Implementation strategies included training, creation of measurement frameworks, implementation of data collection tools, academic support, and emergence of evaluation champions. Evaluation data included administrative records and a staff survey. The Reach Effectiveness Adoption Implementation Maintenance framework guided evaluation. Results indicated that participants reported high awareness, beliefs, attitudes, knowledge, and skills in relation to monitoring, evaluation, and learning at the post-training/refinement period timepoint. However, several barriers limited implementation: reduced workforce capacity, prioritisation of client-driven or other work, leave or staff changes, additional responsibilities, waxing and waning engagement, wider program teams being unclear on the process and value of evaluation, and organisational barriers. These barriers led to a divergence from the planned outcome measures. Enablers to this work included physical presence to facilitate informal discussion, flexibility and openness to change, and communication to support staff. Learnings include the need to strengthen organisational evaluation culture, scaling-back the number of outcomes/data collection sources, and greater involvement with the wider staff cohort. These learnings provide a valuable foundation for other place-based community service organisations to implement evaluation capacity building strategies.
Management measures for the mitigation of spray drift of very fine droplets sprayed by a spraying robot
Motivation and physical activity across Chinese adolescents: Based on latent profile analysis
Despite known links between motivation and physical activity, latent profiles of motivation among Chinese adolescents remain unexamined. Using the person-centered approach, this paper explores the relationship between adolescent motivation and physical activity. We aim to identify the latent motivation profiles and examine how these profiles differential predict physical activity (PA) levels, with attention to gender and age variations. This study recruited 571 adolescents (M age = 11.995, SD = 1.519) in southern China by the scale of Sport Motivation Scale and International Physical Activity Questionnaire-Short Form. We conducted latent profile analysis (LPA) to classify motivation subgroups using Mplus. MANOVA and ANOVA were employed to compare PA differences across profiles, genders, and education levels. The results indicate that three profile model is the optimal model: Low Motivation-High Amotivation (8.45%), Moderate Motivation-High Amotivation (60.61%), and High Autonomous Motivation (30.94%). The subgroup with higher scores of intrinsic motivation and external motivation reported more PA. Moreover, male’s PA is significantly more active than female adolescents, while older adolescents have less PA than younger adolescents. This study identified adolescents with different motivation profiles and PA. Findings suggest the need for more personalized strategies to promote adolescent participation in PA and provide a novel insight into intervention for adolescents with low motivation. Further research could be measured by objective methods and long-term design.
Dysregulation of adipocytokines via the hsa-miR-548ay-3p/PPARγ signaling pathway leads to cardiac fibrosis
A privacy preserving machine learning framework for medical image analysis using quantized fully connected neural networks with TFHE based inference
Presaccadic attentional shifts are not modulated by saccade amplitude
Abstract Humans constantly explore the visual environment through saccades, bringing relevant visual stimuli to the center of the gaze. Before the eyes begin to move, visual attention is directed to the intended saccade target. As a consequence of this presaccadic shift of attention (PSA), visual perception is enhanced at the future gaze position. PSA has been investigated in a variety of saccade amplitudes, from microsaccades to locations that exceed the oculomotor range. Interestingly, recent studies have shown that PSA effects on visual perception are not equally distributed around the visual field. However, it remains unknown whether the magnitude of presaccadic perceptual enhancement varies with the amplitude of the saccades. Here, we measured contrast sensitivity thresholds during saccade planning in a two-alternative forced-choice (2AFC) discrimination task in human observers. Filtered pink noise (1/f) patches, presented at four eccentricities scaled in size according to the cortical magnification factor were used as visual targets. This method was adopted to mitigate well-known eccentricity effects on perception, thereby enabling us to explore the effects associated to saccade amplitudes. First, our results show that saccade preparation enhanced contrast sensitivity in all tested eccentricities. Importantly, we found that this presaccadic perceptual enhancement was not modulated by the amplitude of the saccades. These findings suggest that presaccadic attention operates consistently across different saccade amplitudes, enhancing visual processing at intended gaze positions regardless of saccade size.