Background:Accumulating evidence and medical guidelines recommend high-dose neurorehabilitation for recovery after stroke. The reality, however, is that most patients receive a fraction of this dose, with therapist availability and costs of delivery being major implementational barriers. Objective:This study aimed to explore a potential solution by conducting a retrospective analysis of a real-world enhanced clinical service that used gamified self-training technologies at home under remote therapist supervision. Methods:Data from 17 patients who completed a 12-18 week full-body, high-dose neurorehabilitation program entirely at home were analyzed. Program delivery relied primarily on patients training independently (asynchronously) with the MindMotion GO (MindMaze) gamified-therapy solution. Accompanying telerehabilitation training sessions with a therapist occurred weekly while therapists used a web application to continuously monitor and manage the program remotely. Effectiveness of the program was assessed through measured active training time, a measure that more closely reflects delivered dosage as opposed to scheduled dose. Patient recovery was evaluated with standardized impairment and functional clinical measures and patient self-reported outcome measures. Finally, a cost model was computed to evaluate the resource efficiency of the program. Results:Patients maintained high training adherence throughout the program and reached an average total active training time of 39.7 (SD 21.4) hours, with the majority delivered asynchronously (mean 82.2%, SD 10.8%). Patients improved in both upper-limb (Fugl-Meyer Upper Extremity, mean 6.4, SD 5.1; P<.001) and gait and balance measures (Functional Gait Assessment, mean 3.1, SD 2.6; P<.001; Berg Balance Scale, mean 6.1, SD 4.4; P<.001). Overall, the program was viewed very favorably among patients who completed a post-program survey, with 73.7% (14/19) of respondents being satisfied or very satisfied, while 63.2% (12/19) of respondents reported subjective improvements in physical abilities. Per-patient therapist costs approximated US $338, representing a resource-efficient alternative to delivering the same dose via one-on-one in-person training sessions (US $1903). Conclusions:This work demonstrates effective high-dose neurorehabilitation delivery via gamified therapy technologies at home. The approach shows that training time can be successfully decoupled from therapist-presence without compromising adherence, outcomes, or patient satisfaction over an extended program period. Given growing concerns over therapist availability and increasing health care costs, this resource-efficient approach can help achieve medical guidelines and complement existing clinic-based approaches.

The introduction of 5G networks has significantly advanced communication technology, offering faster speeds, lower latency, and greater capacity. This progress sets the stage for Beyond 5G (B5G) networks, which present new complexity and performance requirements challenges. Linear Programming (LP), Integer Linear Programming (ILP), and Mixed-Integer Linear Programming (MILP) models have been widely used to model the optimization of resource allocation problems in networks. This paper reviews 103 studies on resource allocation strategies in 5G and B5G, focusing specifically on optimization problems modelled as LP, ILP, and MILP. The selected studies are categorized based on network architectures, types of resource allocation problems, and specific objective functions and constraints. The review also discusses solution methods for NP-hard ILP and MILP problems by categorizing the solution methods into different categories. Additionally, emerging trends, such as integrating AI and machine learning with optimization models, are explored, suggesting promising future research directions in network optimization. The paper concludes that LP, ILP, and MILP models have been widely adopted across various network architectures, resource types, objective functions, and constraints and remain critical to optimizing next-generation networks.

Detecting uneven and imperfect road segments is crucial for self-driving vehicles to ensure a safe and comfortable autonomous driving experience. This paper aims to explore the PointNet architecture across its many evolutions and examine how each generation performs when applied to road irregularity point cloud segmentation. The architectural differences between PointNet, PointNet++, and PointNeXt are explored, highlighting how each improves upon its predecessor. Customized versions of PointNet and PointNeXt are then developed and discussed, each obtaining overall training accuracies of 91% and 99.1%, respectively. These tailored implementations of PointNet and PointNeXt are then compared with an existing PointNet++ model for road unevenness segmentation, with our custom PointNeXt model demonstrating a dramatic 76% decrease in model size and a 0.7% jump in accuracy. Our findings highlight the significant improvements across PointNet generations, especially when used for LiDAR point cloud segmentation tasks. Moreover, this study demonstrates the superiority of PointNeXt for real-time road unevenness detection, opening avenues for further research into better autonomous driving solutions.

Automatic movie genre detection is vital for improving content recommendations, user experiences, and organization. Multi-label generation detection assigns multiple labels to a movie and recognizes a movie’s diverse themes. Although there are many existing methods for generating multiple genre labels from movies but do not provide comprehensive analysis and visual depiction. This work introduces GenVis, a visualization system that provides a better understanding of multi-label genres extracted from movie trailers. The system initially uses text and visual features to classify trailers and assign multiple genre labels and probabilities. Next, GenVis provides four visualization views: a video view for trailer observation, an overall genre view for getting insights into genre distribution, a genre timeline view for temporal genre evolution, and finally, a genre flow summary for more focused genre analysis. The system allows users to pause the frames, sort the results, and process multiple videos. The multi-label classification is rigorously evaluated using MSE, cross-entropy loss, precision, recall, F1-score metrics, achieving high accuracy, and demonstrating strong genre correlations with notable precision in effectively classifying and distinguishing movie genres. Additionally, a user evaluation for visualization evaluation demonstrated the effectiveness and intuitive usability of GenVis with a high overall rating of 4.25 out of 5.0.

Automatic cyberbullying detection in social media is increasingly vital due to the integral role of social networks in people’s lives and the severe impact of cyberbullying. Cyberbullying involves intentional, repetitive, aggressive behaviour to harm others online. Among Urdu-speaking communities worldwide, it is common to use Urdu, Roman Urdu, and English in social media conversations. Existing research and detection methods overlook these linguistic dynamics and fail to address cyberbullying across these languages comprehensively. Additionally, there is no dataset in Urdu and Roman Urdu covering the repetition and intent to harm components of cyberbullying. This research addresses this gap by developing and annotating a comprehensive dataset capturing linguistic variations in cyberbullying instances across Urdu, Roman Urdu, and English, incorporating all aspects of cyberbullying. Besides proposing a dataset, a framework for detecting cyberbullying has been proposed. The framework classifies text messages as aggressive or non-aggressive and introduces novel quantitative measures for repetition and the level of intent to cause harm. The proposed framework classifies cyberbullying by applying thresholds to measures of aggression, repetition, and intent to harm, integrating all three aspects. Results show aggression detection using fine-tuned m-BERT and MuRIL, incorporating measures of repetition and intent to harm on the proposed dataset. Additionally, experiments are conducted to demonstrate the impact of repetition and intent to harm on cyberbullying classification. The best results on the dataset are achieved using fine-tuned MuRIL with a precision of 0.93, recall of 0.92, and an F-measure of 0.92 by incorporating quantitative measures of repetition and intent to harm.

Tertiary studies are conducted to offer a comprehensive perspective on a subject by compiling secondary literature at a meta-level. This study appraises secondary studies in computer vision applications for infrastructure management using drone-captured imagery to investigate different dimensions, trends and quality of secondary studies. This tertiary study uses three databases to select studies published from 2018 to 2023. A total of 57 secondary studies are analyzed. Various demographic and temporal patterns are examined by assessing the prevalence of secondary studies concerning the year of publication, publishing platforms, and the nature of the synthesis carried out. The quality of the secondary studies is evaluated using the Database of Abstracts of Reviews of Effects (DARE) criteria. The thematic analysis identifies six major application areas in infrastructure management, with miscellaneous applications categorized separately. The findings of the study offer a comprehensive overview of technological advancements, challenges, and potential applications in infrastructure management using drone imagery.

Task-specific dystonia leads to loss of sensorimotor control for a particular motor skill. Although focal in nature, it is hugely disabling and can terminate professional careers in musicians. Biomarkers for underlying mechanism and severity are much needed. In this study, we designed a keyboard device that measured the forces generated at all fingertips during individual finger presses. By reliably quantifying overflow to other fingers in the instructed (enslaving) and contralateral hand (mirroring) we explored whether this task could differentiate between musicians with and without dystonia. 20 right-handed professional musicians (11 with dystonia) generated isometric flexion forces with the instructed finger to match 25%, 50% or 75% of maximal voluntary contraction for that finger. Enslaving was estimated as a linear slope of the forces applied across all instructed/uninstructed finger combinations. Musicians with dystonia had a small but robust loss of finger dexterity. There was increased enslaving and mirroring, primarily during use of the symptomatic hand (enslaving p = 0.003; mirroring p = 0.016), and to a lesser extent with the asymptomatic hand (enslaving p = 0.052; mirroring p = 0.062). Increased enslaving and mirroring were seen across all combinations of finger pairs. In addition, enslaving was exaggerated across symptomatic fingers when more than one finger was clinically affected. Task-specific dystonia therefore appears to express along a gradient, most severe in the affected skill with subtle and general motor control dysfunction in the background. Recognition of this provides a more nuanced understanding of the sensorimotor control deficits at play and can inform therapeutic options for this highly disabling disorder.

The increasing usage of social media networks has raised concerns about the growing frequency of cyberbullying incidents. The definition of cyberbullying lacks universal consensus, yet according to several authors, cyberbullying is characterized by aggressive, repetitive, and intentional communication among peers. However, existing cyberbullying detection datasets often focus solely on classifying texts as aggressive or non-aggressive, neglecting the other cyberbullying aspects, thus hindering research progress. This paper proposes a framework for designing a new dataset incorporating all four aspects of cyberbullying to address this gap. The text messages are sourced from a real dataset, while the users’ data is generated synthetically. The resulting dataset contains messages exchanged randomly among different pairs of users, thus inculcating repetition. Additionally, the degree of peerness, defined and calculated to measure the likelihood of two users being peers, is used. The intent of harm is quantified as a numeric value using the ratios of aggression and repetition. As a result, the proposed dataset encompasses all four aspects of cyberbullying by providing repeated aggressive messages among users along with quantitative values of the degree of peerness and intent to harm. The proposed dataset is adaptable, with adjustable threshold values for peerness, repetition, and intent to harm, offering flexibility for various applications. The paper concludes by presenting the results of some baseline machine-learning methods on the proposed dataset.

What happens once a cortical territory becomes functionally redundant? We studied changes in brain function and behavior for the remaining hand in humans (male and female) with either a missing hand from birth (one-handers) or due to amputation. Previous studies reported that amputees, but not one-handers, show increased ipsilateral activity in the somatosensory territory of the missing hand (i.e., remapping). We used a complex finger task to explore whether this observed remapping in amputees involves recruiting more neural resources to support the intact hand to meet greater motor control demands. Using basic fMRI analysis, we found that only amputees had more ipsilateral activity when motor demand increased; however, this did not match any noticeable improvement in their behavioral task performance. More advanced multivariate fMRI analyses showed that amputees had stronger and more typical representation-relative to controls' contralateral hand representation-compared with one-handers. This suggests that in amputees, both hand areas work together more collaboratively, potentially reflecting the intact hand's efference copy. One-handers struggled to learn difficult finger configurations, but this did not translate to differences in univariate or multivariate activity relative to controls. Additional white matter analysis provided conclusive evidence that the structural connectivity between the two hand areas did not vary across groups. Together, our results suggest that enhanced activity in the missing hand territory may not reflect intact hand function. Instead, we suggest that plasticity is more restricted than generally assumed and may depend on the availability of homologous pathways acquired early in life.

Road infrastructure is essential for transportation safety and efficiency. However, the current methods for assessing road conditions, crucial for effective planning and maintenance, suffer from high costs, time-intensive procedures, infrequent data collection, and limited real-time capabilities. This paper presents an efficient lightweight system to analyze road quality from video feeds in real-time. The backbone of the system is EdgeFusionViT, a novel vision transformer (ViT)-based architecture that uses an attention-based late fusion mechanism. The proposed architecture outperforms lightweight CNN-based and ViT-based models. Its practicality is demonstrated by its deployment on an edge device, the Nvidia Jetson Orin Nano, enabling real-time road analysis at 12 frames per second. EdgeFusionViT outperforms existing benchmarks, achieving an impressive accuracy of 89.76% on the Road Surface Condition Dataset (RSCD). Notably, the model maintains a commendable accuracy of 76.89% even when trained with only 2% of the dataset, demonstrating its robustness and efficiency. These findings highlight the system’s potential in road infrastructure management. It aids in creating safer, more efficient transport systems through timely, accurate road condition assessments. The study sets a new benchmark and opens up possibilities for advanced machine learning in infrastructure management.

Surveillance video analysis using automated AI-based techniques is a prominent research field with real-world applications. Several techniques aiming at recognizing activities, behaviour, and violent actions are present in literature. Real-world data analysis for violence detection is still a major challenge due to the limited datasets available for training with complex scenarios and varied scaling of objects performing different activities. In this paper, we present a novel deep learning model considering specially designed frame encoders for spatial feature extraction that are generalized towards many challenges, such as light conditions and indoor and outdoor scenarios. Furthermore, the spatial features in the stacked form are analyzed using a temporal deep learning model to observe and learn the temporal patterns dependencies by considering past and future information while predicting the violent or normal class. In the literature, violence is considered a binary classification of either fight or no fight. Different from these techniques, we present a multi-class classification of violent activities considering different types of human violence, such as assault, shooting, etc. The data for multi-class violence classification is extracted from the famous real-world anomaly detection UCF-crime dataset, where only six human-involved types of anomalies are considered for preparing training and testing sets. We report 41.1

In recent years, the rising use of social media has propelled automated cyberbullying detection into a prominent research domain. However, challenges persist due to the absence of a standardized definition and universally accepted datasets. Many researchers now view cyberbullying as a facet of cyberaggression, encompassing factors like repetition, peer relationships, and harmful intent in addition to online aggression. Acquiring comprehensive data reflective of all cyberbullying components from social media networks proves to be a complex task. This paper provides a description of an extensive semi-synthetic cyberbullying dataset that incorporates all of the essential aspects of cyberbullying, including aggression, repetition, peer relationships, and intent to harm. The method of creating the dataset is succinctly outlined, and a detailed overview of the publicly accessible dataset is additionally presented. This accompanying data article provides an in-depth look at the dataset, increasing transparency and enabling replication. It also aids in a deeper understanding of the data, supporting broader research use.

The heart is one of the vital organs of the human body whose pumping action ensures the supply of blood to every single cell of the body. This supply may be interrupted because of cardiac arrest which directly affects the drainage of blood from the heart to other body parts causing arrhythmias and is one of the underlying causes of cardiovascular death. However, this effect can be reversed by Cardiopulmonary resuscitation which is a life-saving technique. CPR is done on patients manually to revive the blood flow & pumping action of the heart, enabling it to pump effectively. To avoid the potential negative impacts of applying excess pressure on ribs during CPR which may cause rib fracture to the individual, a complete understanding of delivering CPR is required. Having this idea, we design a cost-effective CPR Controller Manikin to train health care providers & nursing students about CPR to improve their efficiency in lesser time. This training tool can take account of all the basic parameters of Quality CPR(QCPR) including no. of compressions, force delivered, depth of compressions, no. of breath, correct hand position, and air pressure delivered & displaying them on software. The gadget is distinctive as it not only aids in the training of the health care providers but also ensures that the procedure is up to the mark in delivering QCPR to the patient

The study of automated video surveillance systems study using computer vision techniques is a hot research topic and has been deployed in many real-world CCTV environments. The main focus of the current systems is higher accuracy, while the assistance of surveillance experts in effective data analysis and instant decision making using efficient computer vision algorithms need researchers' attentions. In this research, to the best of our knowledge, we are the first to introduce a process control technique: control charts for surveillance video data analysis. The control charts concept is merged with a novel deep learning-based violence detection framework. Different from the existing methods, the proposed technique considers the importance of spatial information, as well as temporal representations of the input video data, to detect human violence. The spatial information are fused with the temporal dimension of the deep learning model using a multi-scale strategy to ensure that the temporal information are properly assisted by the spatial representations at multi-levels. The proposed frameworks' results are kept in the history-maintaining module of the control charts to validate the level of risks involved in the live input surveillance video. The detailed experimental results over the existing datasets and the real-world video data demonstrate that the proposed approach is a prominent solution towards automated surveillance with the pre- and post-analyses of violent events.

The Big Video Data generated in today's smart cities has raised concerns from its purposeful usage perspective, where surveillance cameras, among many others are the most prominent resources to contribute to the huge volumes of data, making its automated analysis a difficult task in terms of computation and preciseness. Violence detection (VD), broadly plunging under action and activity recognition domain, is used to analyze Big Video data for anomalous actions incurred due to humans. The VD literature is traditionally based on manually engineered features, though advancements to deep learning based standalone models are developed for real-time VD analysis. This paper focuses on overview of deep sequence learning approaches along with localization strategies of the detected violence. This overview also dives into the initial image processing and machine learning-based VD literature and their possible advantages such as efficiency against the current complex models. Furthermore,the datasets are discussed, to provide an analysis of the current models, explaining their pros and cons with future directions in VD domain derived from an in-depth analysis of the previous methods.

Musician's dystonia presents with a persistent deterioration of motor control during musical performance. A predominant hypothesis has been that this is underpinned by maladaptive neural changes to the somatotopic organization of finger representations within primary somatosensory cortex. Here, we tested this hypothesis by investigating the finger-specific activity patterns in the primary somatosensory and motor cortex using functional MRI and multivariate pattern analysis in nine musicians with dystonia and nine healthy musicians. A purpose-built keyboard device allowed characterization of activity patterns elicited during passive extension and active finger presses of individual fingers. We analysed the data using both traditional spatial analysis and state-of-the art multivariate analyses. Our analysis reveals that digit representations in musicians were poorly captured by spatial analyses. An optimized spatial metric found clear somatotopy but no difference in the spatial geometry between fingers with dystonia. Representational similarity analysis was confirmed as a more reliable technique than all spatial metrics evaluated. Significantly, the dissimilarity architecture was equivalent for musicians with and without dystonia. No expansion or spatial shift of digit representation maps were found in the symptomatic group. Our results therefore indicate that the neural representation of generic finger maps in primary sensorimotor cortex is intact in musician's dystonia. These results speak against the idea that task-specific dystonia is associated with a distorted hand somatotopy and lend weight to an alternative hypothesis that task-specific dystonia is due to a higher-order disruption of skill encoding. Such a formulation can better explain the task-specific deficit and offers alternative inroads for therapeutic interventions.

Musicians’ dystonia presents with a persistent loss of motor control during musical performance. The pre- dominant hypothesis is that this loss of motor control is underpinned by maladaptive neural changes to the somatotopic organization of finger representations in primary somatosensory cortex. Here, we tested this hypothesis by investigating the finger-specific activity patterns in the primary somatosensory (S1) and motor cortex (M1) using functional magnetic resonance imaging with state-of-the art multivariate analyses in 11 musicians with dystonia and 9 healthy musicians. We also characterized their dexterous finger control to investigate whether the deficit is strictly limited to musical performance or also generalizes to a non-musical task. We report two key findings. First, during the production of individuated finger presses, musicians with dystonia showed a small, but robust loss of motor control. This deficit was characterized by both a reduction in finger individuation ability, and an exaggeration of mirror movements primarily during use of the clinically identified symptomatic hand, but also to a lesser extent during asymptomatic hand use. Second, we found no evidence of disease-related changes in the corresponding finger representa- tions in S1/M1. Our results contradict the view that abnormalities in sensorimotor finger representations play a role in the pathophysiology of musicians’ dystonia. Our behavioral results also suggest that the loss of finger dexterity in musicians’ dystonia expresses along a spectrum with subtle abnormalities in motor control evident during ordinary dexterous tasks. asadnick@sgul.ac.uk|ABN Bursary 39

Abstract In software, code is the only part that remains up to date, which shows how important code is. Code readability is the capability of the code that makes it readable and understandable for professionals. The readability of code has been a great concern for programmers and other technical people in development team because it can have a great influence on software maintenance. A lot of research has been done to measure the influence of program constructs on the code readability but none has placed the highly influential constructs together to predict the readability of a code snippet. In this article, we propose a novel framework using statistical modeling that extracts important features from the code that can help in estimating its readability. Besides that using multiple correlation analysis, our proposed approach can measure dependencies among di erent program constructs. In addition, a multiple regression equation is proposed to predict the code readability. We have automated the proposals in a tool that can do the aforementioned estimations on the input code. Using those tools we have conducted various experiments. The results show that the calculated estimations match with the original values that show the effectiveness of our proposed work. Finally, the results of the experiments are analyzed through statistical analysis in SPSS tool to show their significance.

It has been proposed that a form of cortical reorganization (changes in functional connectivity between brain areas) can be assessed with resting-state (rs) functional MRI (fMRI). Here, we report a longitudinal data set collected from 19 patients with subcortical stroke and 11 controls. Patients were imaged up to five times over 1 year. We found no evidence, using rs-fMRI, for longitudinal poststroke cortical connectivity changes despite substantial behavioral recovery. These results could be construed as questioning the value of resting-state imaging. Here, we argue instead that they are consistent with other emerging reasons to challenge the idea of motor-recovery-related cortical reorganization poststroke when conceived of as changes in connectivity between cortical areas. NEW & NOTEWORTHY We investigated longitudinal changes in functional connectivity after stroke. Despite substantial motor recovery, we found no differences in functional connectivity patterns between patients and controls, nor any changes over time. Assuming that rs-fMRI is an adequate method to capture connectivity changes between cortical regions after brain injury, these results provide reason to doubt that changes in cortico-cortical connectivity are the relevant mechanism for promoting motor recovery.

Despite significant infrastructure improvements, cloud computing still faces numerous challenges in terms of load balancing. Several techniques have been applied in the literature to improve load balancing efficiency. Recent research manifested that load balancing techniques based on metaheuristics provide better solutions for proper scheduling and allocation of resources in the cloud. However, most of the existing approaches consider only a single or few QoS metrics and ignore many important factors. The performance efficiency of these approaches is further enhanced by merging with machine learning techniques. These approaches combine the relative benefits of load balancing algorithm backed up by powerful machine learning models such as Support Vector Machines (SVM). In the cloud, data exists in huge volume and variety that requires extensive computations for its accessibility, and hence performance efficiency is a major concern. To address such concerns, we propose a load balancing algorithm, namely, Data Files Type Formatting (DFTF) that utilizes a modified version of Cat Swarm Optimization (CSO) along with SVM. First, the proposed system classifies data in the cloud from diverse sources into various types, such as text, images, video, and audio using one to many types of SVM classifiers. Then, the data is input to the modified load balancing algorithm CSO that efficiently distributes the load on VMs. Simulation results compared to existing approaches showed an improved performance in terms of throughput (7%), the response time (8.2%), migration time (13%), energy consumption (8.5%), optimization time (9.7%), overhead time (6.2%), SLA violation (8.9%), and average execution time (9%). These results outperformed some of the existing baselines used in this research such as CBSMKC, FSALB, PSO-BOOST, IACSO-SVM, CSO-DA, and GA-ACO.