Federated learning trains a neural network model using the client's data to maintain the benefits of centralized model training, while maintaining the privacy. However, if the client data are not independently and identically distributed (non-IID) because of different environments, the accuracy of the model may suffer from client drift during training owing to discrepancies in each client's data. This study proposes a personalized federated learning algorithm based on the concept of multitask learning to divide each client model into two layers: a feature extraction layer and a category prediction layer. The feature extraction layer maps the input data to a low-dimensional feature vector space. Furthermore, the parameters of the neural network are aggregated with those of other clients using an adaptive method. The category prediction layer maps low-dimensional feature vectors to the label sample space, with its parameters remaining unaffected by other clients to maintain client uniqueness. The proposed personalized federated learning method produces faster learning model convergence rates and higher accuracy rates for the non-IID datasets in our experiments.

Optimization is critical in various fields like smart vehicles, and transportation. Federated Learning (FL) has emerged as an effective approach in the coordination of autonomous vehicles, but traditional methods such as FedAvg can create performance disparities across clients. This paper addresses this fairness issue through the $q$-Fair Federated Learning ($q$-FFL) framework, adjusting model performance across clients using a tunable fairness parameter. We propose a modified FedAvg algorithm for $q$-FFL that maintains comparable convergence rates, ensuring more balanced client outcomes. Additionally, we explore incentive mechanisms in FL using a Stackelberg game model, incorporating a fairness coefficient to encourage equitable participation. Building on prior works, we redefine client utility functions to address communication and computation costs, ensuring fair resource allocation. The proposed framework achieves both global and local fairness, maintaining a unique Nash equilibrium in the modified game setting.

Federated learning is an effective approach for preserving data privacy and security, enabling machine learning to occur in a distributed environment and promoting its development. However, an urgent problem that needs to be addressed is how to encourage active client participation in federated learning. The Shapley Value, a classical concept in cooperative game theory, has been utilized for data valuation in machine learning services. Nevertheless, existing numerical evaluation schemes based on the Shapley Value are impractical as they necessitate additional model training, leading to increased communication overhead. Moreover, participants’ data may exhibit Non-IID characteristics, posing a significant challenge to evaluating participant contributions. Non-IID data weakens the marginal effects of participants, which in turn affects model accuracy and leads to underestimated contribution measurements. Current works often overlook the impact of heterogeneity on model aggregation. This paper presents a fair federated learning contribution measurement scheme that addresses the need for additional model computations. By introducing a noval aggregation weight, it enhances the accuracy of contribution measurement. Experiments on the MNIST dataset show that the proposed method can accurately compute the contributions of participants. Compared to existing baseline algorithms, the model accuracy is significantly improved, with a similar time cost.

Data security and user privacy concerns are increasingly gaining attention. Federated learning models based on differential privacy offer a distributed machine learning framework that protects data privacy; however, the added noise can impact the model's utility, making performance evaluation crucial. To optimize the balance between privacy protection and model performance, we propose the Adaptive Weight-Based Differential Privacy Federated Learning (AWDP-FL) framework. This framework processes model parameters from the perspectives of neural network layers and model weights. Initially, each participant trains the model locally. During iterative training, the clipping threshold is determined by selecting an adaptive weight coefficient, followed by adaptive gradient clipping to control the gradient magnitude. Subsequently, adaptive gradient updates are applied to the model, and dynamic Gaussian noise is introduced when uploading the parameters to protect participant privacy. The server aggregates these noise-perturbed parameters to update the global model. This framework ensures strong privacy protection while maintaining model accuracy. Theoretical analysis and experimental results validate the effectiveness of this framework under stringent privacy constraints.

In the industrial sector, intelligent sensors in fault diagnosis are becoming more critical in recent technological improvements. The prediction accuracy can be enhanced in fault diagnosis using 3-dimensional, sequential, real-time, and image data. Sensors that capture the vibration, sound, and image data are more critical in predicting unbalancing, tool wear, crack, misalignment, etc, in the rotating machinery to increase productivity and to provide an effective maintenance management system. Due to the fast development of industry 4.0 techniques, monitoring of mechanical machinery is experiencing explosive growth and getting more attention in the area of fault diagnosis (FD). Machine learning and Deep learning methods give promising results and accuracy in predicting faults in rotating machinery on shop floors. The success of AI-based models is due to the availability of comprehensive labeled data. Federated Learning (FL) is the machine learning subfield aiming to train an algorithm with a heterogeneous dataset. Data transmission from local facilities to a central server in AI models creates data privacy and security issues. Heterogeneous data analysis is a complicated process in predicting the machine fault in the central server because of millions of data during real-time condition monitoring. Decentralized data handling and analysis is mandatory in condition monitoring because of heterogeneity in data processing, data privacy, and security advantages. The application of federated learning in fault diagnosis has been getting more attention in recent days, and this study is a review of FL applications that address fault diagnosis in rotating machinery in the first phase. A comparison between the types of FL approaches in FD and the use of aggregation algorithms and their applications will also be discussed in Phase 1. In phase 2, a novel methodology has been proposed using Federated learning to diagnose rotating machinery faults. The proposed method, FLOACOS, addresses how prediction challenges are solved using federated learning approaches by optimizing and standardizing the data at local facilities. This work will be helpful for future condition monitoring researchers and gives an overview of a novel method of the FL technique used in predicting faults and the progress made in the maintenance management of rotating machinery.

Federated learning (FL) has become a burgeoning and attractive research area, which provides a collaborative training scheme for distributed data sources with privacy concerns. Most existing FL studies focus on taking unimodal data, such as images and text, as the model input and resolving the heterogeneity challenge, i.e., the non-identically distributed (non-IID) challenge caused by data distribution imbalance related to data labels and data amount. In real-world applications, data are usually described by multiple modalities. However, to the best of our knowledge, only a handful of studies have been proposed to improve the system performance by utilizing multimodal data. In this survey paper, we identify the significance of this emerging research topic – multimodal federated learning (MFL) and perform a literature review on the state-of-art MFL methods. Furthermore, we categorize multi-modal federated learning into congruent and incongruent multimodal federated learning based on whether all clients possess the same modal combinations. We investigate the feasible application tasks and related benchmarks for MFL. Lastly, we summarize the promising directions and fundamental challenges in this field for future research.

The Internet of Things (IoT) has revolutionized various industries, but the increased dependence on all kinds of IoT devices and the sensitive nature of the data accumulated by them pose a formidable threat to privacy and security. While traditional Intrusion Detection Systems (IDSs) have been effective in securing critical infrastructures, the centralized nature of these systems raises serious data privacy concerns, as sensitive information is sent to a central server for analysis. This research paper introduces a Federated Learning (FL) approach designed for detecting intrusions in diverse IoT networks that addresses the issue of data privacy by ensuring that sensitive information is kept in the individual IoT devices during model training. Our framework utilizes the Federated Averaging (FedAvg) algorithm, which aggregates model weights from distributed devices to refine the global model iteratively. The proposed model manages to achieve above 90\% accuracy across various metrics, including precision, recall, and f1-score, while maintaining low computational demands. The results show that the proposed system successfully identifies various types of cyber attacks, including DoS, DDoS, Data Injection, Ransomware, and several others showcasing its robustness. This research makes a great advancement to the intrusion detection systems by providing an efficient and reliable solution that is more scalable and privacy friendly than any of the existing models.

Federated learning techniques aim to train and build machine learning models based on distributed datasets across multiple devices, avoiding data leakage. The main idea is to perform training on remote devices or isolated data centers without transferring data to centralized repositories, thus mitigating privacy risks. Data analytics in education, in particular learning analytics, is a promising scenario to apply this approach to address the legal and ethical issues related to processing sensitive data. Indeed, given the nature of the data to be studied (personal data, educational outcomes, data concerning minors), it is essential to ensure that the conduct of these studies and the publication of the results provide the necessary guarantees to protect the privacy of the individuals involved and the protection of their data. In addition, the application of quantitative techniques based on the exploitation of data on the use of educational platforms, student performance, use of devices, etc., can account for educational problems such as the determination of user profiles, personalized learning trajectories, or early dropout indicators and alerts, among others. This paper presents the application of federated learning techniques to two learning analytics problems: dropout prediction and unsupervised student classification. The experiments allow us to conclude that the proposed solutions achieve comparable results from the performance point of view with the centralized versions, avoiding centralizing the data for training the models.

This paper systematically discusses the application and development of federated learning in data privacy protection and data value sharing. With the rapid development of global information technology, especially the explosive growth of data from Internet of Things devices, data security and privacy protection are facing unprecedented challenges. This paper first analyzes the growth trend of global data volume and its importance to next generation technologies such as artificial intelligence technologies such as deep learning. Second, the paper provides an in-depth look at the impact of current data privacy regulations on data flows and value creation, particularly the EU's GDPR and China's Data Security and Personal Information Protection Law. Then, this paper introduces in detail federated learning, as a new distributed machine learning paradigm, which effectively solves the contradiction between existing data sharing and privacy protection by protecting individual data privacy and realizing global model collaborative construction. Finally, this paper discusses the combination of blockchain technology and federated learning, and proposes BeFL architecture as a new secure, decentralized and trusted federated learning system, which is expected to provide a comprehensive solution for large-scale data processing and value creation in multi-party scenarios. The research in this paper not only deepens the understanding of federation learning in theory, but also provides important reference and enlightenment for future research and application in related fields.

Traditional classroom teaching model can no longer meet the needs of students' ability development. As a new model actively advocated in the modern education concept, personalized teaching focuses on the development needs of students and emphasizes respect for students' individualized growth. In this work, we utilize deep learning method to recommend teaching strategies for different features of students. Specifically, the proposed method includes the construction of a multi-layer neural network model, which can analyze students' multi-dimensional learning data and behavior patterns, so as to recommend the most suitable teaching strategies for individual students. Through data collection and preprocessing, model architecture design, model training and optimization, the model gradually improves the accuracy of prediction and recommendation. Experimental analysis verifies the effectiveness of the model through field teaching experiments, and the results show that the personalized teaching strategy based on deep learning can significantly improve the learning effect and participation of students.

This paper addresses the evolving landscape of electricity markets in Europe, with a focus on the integration of Renewable Energy Communities as introduced by the Renewable Energy Directive 2018/2001. Residents within a postal code area are highly incentivized to join a community, which enables them to exchange energy among themselves at lower procurement costs. Thereby, energy management systems optimize the operation of respective energy systems, with electrical load forecasting playing a key role. Given that prosumers may switch between communities on a daily basis, electricity demands within these groups will vary, leading to data that is non-stationary, discontinuous as well as non-identical and independently distributed. To encounter this issue, we propose a sophisticated forecasting model that applies federated learning, using informations from various distributed communities to learn domain-invariant features. To achieve this, we initially utilize synthetic electrical load time series at district level and aggregate them to profiles of Renewable Energy Communities with dynamic portfolios. Subsequently, we develop a forecasting model that accounts for the composition of residents of a Renewable Energy Community, adapt data pre-processing in accordance with the time series process, and detail a federated learning algorithm that incorporates weight averaging and data sharing. Following the training of various experimental setups, we ultimately evaluate their effectiveness. The findings suggest that our proposed framework is capable of effectively forecast non-stationary and discontinuous time series and that it can be applied to new, unseen data through the integration of knowledge from multiple sources.

Background. The Stanford Health Assessment Questionnaire-Disability Index (HAQ) is widely used to measure functional ability in persons with Rheumatoid Arthritis (RA). The instrument was developed with limited involvement from persons with RA, and their perception of the instrument has not been studied in depth. The aim of this study was to explore how persons with RA experience the use of the HAQ in care. Methods. Forty persons with RA were purposefully recruited to participate in semi-structured interviews. The interviews were then analyzed qualitatively using thematic analysis. Results. The participants questioned the relevance of the HAQ but nevertheless experienced that the instrument had a profound effect on their understanding of health and how care is delivered. The analysis resulted in three themes: Problems with individual items, meaning of the summative score, and effects on care and health perceptions. Conclusions. To make the HAQ relevant to persons with RA, it needs to be revised or to include an option to select items most meaningful to the respondent. To ensure relevance, the HAQ update should preferably be co-created by researchers, clinicians and persons with RA.

With the widespread application of artificial intelligence technology in various industries, users' attention to privacy and data security has increased significantly. Federated learning, as a new technology paradigm combining privacy-enhanced computing and artificial intelligence, resolves the contradiction between data security and open sharing. This paper presents the benefits of federated learning in terms of privacy, real-time processing, model robustness, compliance and cross-industry applications. At the same time, when combined with Edge AI technology, federated learning promotes the decentralisation of intelligent systems, improving data privacy protection and model accuracy. This paper also discusses the application cases of federated learning in the medical field, through local data processing and model training, effectively protecting user privacy, realizing medical data sharing and model optimization, and promoting the development of artificial intelligence.

The present grievous situation of the tuberculosis disease can be improved by efficient case management and timely follow-up evaluations. With the advent of digital technology this can be achieved by quick summarization of the patient-centric data. The aim of our study was to assess the effectiveness of the ODK Scan paper-to-digital system during testing period of three months. A sequential, explanatory mixed-method research approach was employed to elucidate technology use. Training, smartphones, application and 3G enabled SIMs were provided to the four field workers. At the beginning, baseline measures of the data management aspects were recorded and compared with endline measures to see the impact of ODK Scan. Additionally, at the end, users’ feedback was collected regarding app usability, user interface design and workflow changes. 122 patients’ records were retrieved from the server and analysed for quality. It was found that ODK Scan recognized 99.2% of multiple choice bubble responses and 79.4% of numerical digit responses correctly. However, the overall quality of the digital data was decreased in comparison to manually entered data. Using ODK Scan, a significant time reduction is observed in data aggregation and data transfer activities, however, data verification and form filling activities took more time. Interviews revealed that field workers saw value in using ODK Scan, however, they were more concerned about the time consuming aspects of the use of ODK Scan. Therefore, it is concluded that minimal disturbance in the existing workflow, continuous feedback and value additions are the important considerations for the implementing organization to ensure technology adoption and workflow improvements.

In today’s Web 2.0 era, online social media has become an integral part of our lives. In the course of the information revolution, the form of information has undergone a radical change, from simple text information to today’s integrated video, image, text and audio, and there has also been a great change in the way of dissemination and access, as people nowadays do not just rely on traditional media to passively receive information, but more actively and selectively obtain information from social media. Therefore, it has become a great challenge for us to effectively utilize these massive and integrated multi-modal media information to form an effective system of retrieval, browsing, analysis and usage. Unlike movies and traditional long-form video content, micro-videos are usually short in length, between a few seconds and tens of seconds, which allows users to quickly browse different contents and make full use of the fragmented time in their lives, while users can also share their micro-videos to their friends or the public, forming a unique social way. Video contains rich multimodal information, and fusing information from multiple modalities in a video recommendation task can improve the accuracy of the video recommendation task.According to the micro-video recommendation task, a new combinatorial network model is proposed to combine the discrete features of each modality into the overall features of various modalities through the network, and then fuse the various modal features to obtain the overall video features, which will be used for recommendation. In order to verify the effectiveness of the algorithm proposed in this paper, experiments are conducted in the public dataset, and it is shown the effectiveness of our model.

Arguably the most important parameter in treating cognitive decline associated with Alzheimer’s disease is the length of time in which improvement, if achieved at all, is sustained. However, monotherapies such as donepezil and memantine are associated with a more rapid decline than no treatment in patients over multi-year follow-ups {Kennedy, 2018 #535}. Furthermore, anti-amyloid antibody treatment, which at best simply slows decline, is associated with accelerated cerebral atrophy [2], resulting in earlier dementia-associated brain volumes for those treated at the MCI stage than untreated patients. In contrast, a precision medicine approach, in which the multiple potential drivers of cognitive decline are identified for each patient and then targeted with a personalized protocol (such as ReCODE), has led to documented improvement in patients with cognitive decline [3]; [4]; [5], but long-term follow-up (>5 years) has not been reported previously. Therefore, we report here sustained cognitive improvement, in some cases for over a decade, in patients treated with a precision medicine protocol—something that has not been reported in patients treated with anti-cholinesterase, glutamate receptor inhibitory, anti-amyloid, or other therapeutic methods. These case studies warrant long-term cohort studies to determine how frequently such sustained cognitive improvements occur in patients treated with precision medicine protocols.

Federated Learning (FL) represents a promising distributed learning methodology, particularly suitable for dynamic and heterogeneous environments characterized by the presence of Internet of Things (IoT) devices and Edge Computing infrastructures. In this context, FL allows you to train machine learning models directly on edge devices, mitigating data privacy concerns and reducing latency due to transmitting data to central servers. However, the heterogeneity of computational resources, the variability of network connections, and the mobility of IoT devices pose significant challenges to the efficient implementation of FL. This work explores advanced techniques for dynamic model adaptation and heterogeneous data management in edge computing scenarios, proposing innovative solutions to improve the robustness and efficiency of federated learning. We present an innovative solution based on Kubernetes which enables the fast application of FL models to Heterogeneous Architectures. Experimental results demonstrate that our proposals can improve the performance of FL in IoT and edge environments, offering new perspectives for the practical implementation of decentralized intelligent systems.

A pandemic can have a huge impact on normal human life and the economy, taking COVID-19 as an example. While the population flow between countries and regions is the main factor affecting the change of a pandemic, exactly as the airline network. Therefore, realizing the overall control of airports is an effective way to control a pandemic. However, restricted to the differences in prevention and control policies in different areas and privacy issues, the patients’ personal data of the medical center cannot be effectively combined with the passengers’ personal data. This prevents more precise airport control decisions from being made. To the end, this paper designs a novel data sharing framework (i.e., PPChain) based on blockchain and federated learning. The experiment shows that the relationship between the epidemic and aircraft transport can be effectively explored by PPChain, without sharing raw data. This approach does not require centralized trust and improves the security of the shareing process.The scheme can help formulate more scientific and rational prevention and control policies on airports’ control. And it can use aerial data to predict pandemics more accurately.

Effective healthcare relies on accurate and timely diagnosis; however, obtaining large amounts of training data while maintaining patient privacy remains challenging. This study introduces a novel approach utilizing federated learning (FL) and a cross-device multi-modal model for clin-ical event classification using vital signs data. Our architecture leverages FL to train machine learning models, including Random Forest, AdaBoost, and SGD ensemble model, on vital signs data from a diverse clientele at a Boston hospital (MIMIC-IV dataset). The FL structure preserves patient privacy by training directly on each client's device without transferring sensitive data. The study demonstrates the potential of FL in privacy-preserving clinical event classification, achieving an impressive accuracy of 98.9%. These findings underscore the significance of FL and cross-device ensemble technology in healthcare applications, enabling the analysis of large amounts of sensitive patient data while safeguarding privacy.

Despite the existing effective treatment methods, tuberculosis (TB) is the second most deadly infectious disease globally, its carriers in the latent and active phases accounting for more than 20% of the world population. An effective method to control TB and reduce the mortality is regular population screening and diagnosis of the latent form of TB in order to take preventive and curative measures. Numerous methods allow diagnosing TB and directly detecting Mycobacterium tuberculosis (M.tb) biomarkers, including M.tb DNA, proteins, and specific metabolites, as well as antibodies produced by the host immune system in response to M.tb. PCR, ELISA, immunofluorescence and immunochemical analyses, flow cytometry, and other methods allow the detection of M.tb biomarkers or the host immune response to M.tb by recording the optical signal of fluorescent or colorimetric dyes included in the diagnostic tools. Current research in biosensors is aimed at increasing the sensitivity of detection, which can be achieved by using brighter and more photostable optical tags containing fluorescent quantum dots. Here, we review current methods for detection of M.tb biomarkers using optical sensor systems, primarily quantum dot–based nanosensors, and summarize M.tb biomarkers whose detection can be made significantly more sensitive by using quantum dot–based nanosensors.

Background: We assessed the pertinence of yearly updating the International Working Group on the Diabetic Foot (IWGDF) risk classification in people with diabetes by quantifying the changes in the risk group and its accuracy in identifying those developing an ulcer (DFU) in a primary care setting. Methods: In our retrospective cohort study, we included all people with diabetes with a foot as-sessment registry between January 2016 and December 2018 in the Baixo Alentejo Local Health Unit. Foot-related data was collected at baseline after one and two years. DFU and/or death until December 2019 were registered. The proportion of people changing their risk status each year was calculated. Accuracy measures of the IWGDF classification to predict DFU occurrence at one, two, and three years were calculated. Results: A total of 2097 people were followed for three years, during which 0.1% died, and 12.4% developed a DFU. After two years, 3.6% of the participants had progressed to a higher-risk group. The IWGDF classification presented specificity values superior to 90% and negative predictive values superior to 99%. Conclusion: Foot risk status can be safely updated every two years instead of yearly. The IWGDF classification can accurately identify those not at risk of DFU.

With the exponential growth of internet usage and mobile devices, the volume of user behavior data has surged, offering significant opportunities and challenges for businesses. This paper explores user behavior analysis in the context of big data, highlighting its pivotal role in driving innovation and strategic decision-making. We begin by discussing the importance of user behavior data and reviewing technological advancements such as machine learning, natural language processing, and data mining that enable deep insights into user behavior. Through a case study on linguistic analysis in livestreaming e-commerce, we demonstrate how text-mining techniques can correlate linguistic characteristics with sales performance, providing actionable marketing insights. We further explore the broader business applications, including personalized recommendations, user portraits, product design, and strategic decision-making. Technical and ethical challenges, such as data integration, privacy concerns, and algorithmic bias, are also addressed. The paper concludes with a discussion on future trends, such as AI and ML advancements, edge computing, IoT proliferation, and ethical AI frameworks, poised to shape user behavior analysis. Our findings underscore the transformative potential of user behavior data in enhancing business operations and user experiences, providing recommendations for future research focused on real-time analytics, privacy-preserving techniques, and ethical implications of data-driven decisions.

Electroencephalography (EEG) signals are the primary source for discriminating the preictal from the interictal stage, enabling early warnings before the seizure onset. The epileptic seizure prediction models face significant challenges due to data scarcity, diversity, and privacy. This paper proposes a three-tier architecture for epileptic seizure prediction associated with the Federated Learning (FL) model that enhances the capability by utilizing the significant amount of seizure patterns from globally distributed patients with data privacy. The determination of the preictal state is influenced by global and local model-assisted decision-making by modeling the two-level edge layer. Integrating the Spiking Encoder (SE) with Graph Convolutional Neural Network (Spiking-GCNN) works as the local model trained using the bi-timescale approach. Each local model utilizes the aggregated seizure knowledge from the different medical centers through FL and determines the preictal probability in the coarse-grained personalization. The Adaptive Neuro-Fuzzy Inference System (ANFIS) is utilized in fine-grained personalization to recognize epileptic seizure patients by examining the outcomes of the FL model, heart rate variability features, and patient-specific clinical features. The proposed seizure prediction is evaluated using benchmark datasets by comparing them with the existing works to demonstrate the potential results.

As a new distributed machine learning (ML) approach, federated learning (FL) shows the great potential to preserve data privacy by enabling distributed data owners to collaboratively build a global model without sharing their raw data. However, the heterogeneity in terms of data distribution and hardware configurations make it hard to select participants from the thousands of nodes. In this paper, we propose a multi-objective node selection approach to improve time-to-accuracy performance while resisting malicious nodes. We firstly design a deep reinforcement learning assisted FL framework. Then the problem of multi-objective node selection under this framework is formulated as a Markov decision process (MDP), which aims to reduce the training time and improve model accuracy simultaneously. Finally, a deep Q-netwok (DQN) based algorithm is proposed to efficiently solve the optimal set of participants for each iteration. Simulation results show that the proposed method not only significantly improves the accuracy and training speed of FL, but has stronger robustness to resist malicious nodes.

Increasingly measured data in the context of smart cities can be used to develop new and innovative business models to increase efficiency and the value of life. A time-series classification algorithm can support to automatize many different processes such as forecasting services. In order to ensure data security and privacy, Federated Learning trains a global model collaboratively on multiple clients. Having different data-distributions and data-quantities across participating clients, neural networks suffer from slow convergence and overfitting. Based on different data-distributions, data-quantities and number of clients, we develop and evaluate different data-clustering strategies to update global model weights in comparison to the state of the art. We use public time-series data, generate various synthetic datasets and train a Relational-Regularized Autoencoder for classification purposes. Our results show an improvement of model performance concerning generalization.

This study examines the contribution of early executive functions (EFs) in the association between fluency in reading and arithmetic. Kindergarten children (N = 1,185) were assessed on Executive Functions skills and on reading and arithmetic fluency in Grade 1 and Grade 3. The analysis revealed that beyond the connection within each domain there is a unidirectional effect between fluency measures, with Grade 1 reading fluency significantly influencing the development of arithmetic fluency in Grade 3. Furthermore, the findings indicate that kindergarten EFs significantly contribute to arithmetic fluency at both time points and to reading fluency in the first grade. Early EF skills also emerged as a mediator of associations between fluency performance in reading and arithmetic, suggesting that the influence of EFs extends beyond individual academic domains. These findings have implications for understanding the cognitive mechanisms that underlie the relations between these academic skills.

This research investigates the application of Fused Filament Fabrication (FFF) in producing WC-10Co hardmetals, focusing binder formulations and understanding its influence on extrusion force and the influence of printing variables on green density of the samples. By examining the interplay between different binder compositions and backbone contents, the study aims to enhance the mechanical properties of the sintered parts while reducing defects inherent in the printing process. Evidence suggested that formulated feedstocks influenced the hardness of the sintered hardmetal, not due to microstructural changes but to macrostructural responses like macro defects introduced during printing, debinding and sintering of the samples. The results demonstrate the critical role of PP-MA content in improving part density and sintered hardness, suggesting the need for specific thermal debinding protocols tailored to each feedstock. The study provides insights into the formulation of feedstocks for hardmetal FFF, proposing a path toward refining manufacturing processes for better quality and performance of 3D printed hardmetal components.

Federated learning (FL) has emerged as a crucial technology in today’s data-centric technological environment, enabling decentralized machine learning and safeguarding user privacy. This study introduces “Federated Operations (FedOps) Mobile,” a novel FL framework optimized for the dynamic and heterogeneous ecosystem of mobile devices. FedOps Mobile enhances traditional FL approaches for mobile devices by integrating real-time operational control and advanced on-device training capabilities using TensorFlow Lite and CoreML, addressing critical challenges in scalability, efficiency, and system heterogeneity. Our approach utilizes a wide range of devices, facilitated by intelligent client-selection mechanisms. This mechanism evaluates the capabilities and readiness of multiple devices per client to ensure fair and efficient network participation. The framework also utilizes remote device control for seamless task management and sustained engagement, enabling continuous learning without compromising the user experience. We conducted extensive experiments to validate the framework’s performance, focusing on three core aspects: operational efficiency, model personalization, and resource optimization in multi-device environments. The results demonstrate that the proposed method is effective for efficient client selection, energy consumption, and model optimization.

The value of landscape, as part of collective heritage, can be acquired by GIS due to the multilayer approach of the spatial configuration. Proficiency in geospatial technologies in order to collect, process, analyze, interpret, visualize and communicate geographic information is being increased by undergraduate and graduate students, but in particular by those who are training to become geography teachers at secondary education. This training can be carried out through personalized learning and distance learning methodology. Personalized GIS education aims to integrate students and enhance their understanding of landscape. Some teaching experiences are shown whereby opportunities offered by WebGIS will be described, through quantitative tools and techniques that will allow this modality of learning and improve its effectiveness. Results of this research show that students, through geospatial technologies, learn landscape as a diversity of elements but also the complexity of physical and human factors involved. Several conclusions will be highlighted: i) the contribution of geospatial training to education for sustainable development; ii) spatial analysis as a mean of skills acquisition about measures for landscape conservation; iii) expanding and applying acquired knowledge to other geographic spaces and different landscapes.

Educational data-mining is an evolving discipline that focuses on the improvement of self-learning and adaptive methods. It is used for finding hidden patterns or intrinsic structures of educational data. In the arena of education, the heterogeneous data is involved and continuously growing in the paradigm of big-data. To extract meaningful information adaptively from big educational data, some specific data mining techniques are needed. This paper presents a clustering approach to partition students into different groups or clusters based on their learning behavior. Furthermore, personalized e-learning system architecture is also presented which detects and responds teaching contents according to the students’ learning capabilities. The primary objective includes the discovery of optimal settings, in which learners can improve their learning capabilities. Moreover, the administration can find essential hidden patterns to bring the effective reforms in the existing system. The clustering methods K-Means, K-Medoids, Density-based Spatial Clustering of Applications with Noise, Agglomerative Hierarchical Cluster Tree and Clustering by Fast Search and Finding of Density Peaks via Heat Diffusion (CFSFDP-HD) are analyzed using educational data mining. It is observed that more robust results can be achieved by the replacement of existing methods with CFSFDP-HD. The data mining techniques are equally effective to analyze the big data to make education systems vigorous.

This study introduces the Data-Driven Personalized Learning Model (DDPLM), a sophisticated framework designed to enhance foreign language acquisition through the integration of big data analytics. Implemented within the educational platforms Edmodo and Duolingo, DDPLM utilizes real-time data processing to tailor learning paths and content dynamically to individual learner needs. Our findings indicate significant improvements in language learning efficiency, engagement, and retention. The model's adaptability across different digital environments showcases its potential scalability and effectiveness in various educational contexts. Additionally, the research addresses the critical role of personalized feedback and adaptive challenges in maintaining learner motivation and promoting deeper linguistic comprehension. The outcomes suggest that DDPLM significantly transforms traditional language education, making it more personalized, efficient, and aligned with individual learning preferences.

Artificial Intelligence (AI) models are expected to have a great impact in the manufacturing industry, optimizing time and resource cost by enabling applications such as predictive maintenance (PM) of production machines. A necessary condition for this is the availability of high quality data collected as close as possible to the process in question. With the advent of 5G equipped multi sensor platforms (MSPs), high sampling rate data can be collected and transmitted for processing in real time. This poses a data security challenge, since this data may give valuable insight into confidential business information of companies. Federated learning (FL) enables the training of AI models with data from multiple sources without it leaving the shop floor, by utilizing distributed computing resources available on premise. This paper introduces an architecture of FL based on data collected from 5G MSPs for enabling PM in industrial environments and discusses its potential benefits and challenges.

The ability to decode words accurately at an early age strongly predicts reading competence later on. Decoding-challenged children may find it difficult to understand texts, which might impede their overall academic development. Strong word decoding skills accelerate the process of learning vocabulary and grammar rules in a variety of languages. The paper aimed to collect relevant literature regarding word decoding ability in the world of the language acquisition process. Policymakers and educators will be greatly impacted by realizing how crucial word decoding is to language learning. Providing targeted interventions for struggling readers, encouraging a positive reading culture, and explicitly teaching phonological and orthographic abilities are all essential components of effective reading education. Promoting decoding proficiency also requires making sure that people have access to excellent reading materials and tools. In order to enable learners to become great readers and communicators across linguistic and cultural boundaries, educators and policymakers should prioritize phonemic awareness from an early age and use effective teaching methodologies.

Employing phonemic awareness in language development can lead to transformative educational outcomes. Having mastered this skill, students have a good foundation in language comprehension, writing, and reading, which increases their chances of academic achievement. This change can affect entire classrooms and schools, fostering a culture of literacy and academic performance, rather than just a select few students. The aim of the study is to gather literature that offers substantial information about this topic based on the facts presented. The overwhelming body of research points to the benefits of teaching phonemic awareness in early childhood education. The foundation of transformative education is phonemic awareness, which fosters fundamental language and reading skills. This profound effect emphasizes how crucial phonemic awareness is to the teaching process. Educators can use phonemic awareness to boost students' language development and create a transformative learning environment by putting these guidelines into practice.

Deep learning dominates automated animal activity recognition (AAR) tasks due to the high performance on large-scale datasets. However, constructing centralised data across diverse farms raises data privacy issues. Federated learning (FL) provides a distributed learning solution to train a shared model by coordinating multiple farms (clients) without sharing their private data. Whereas directly applying FL to AAR tasks often faces two challenges: client-drift during local training and gradient conflicts during global aggregation. In this study, we develop a novel FL framework called FedAAR to achieve AAR with sensor data. Specifically, we devise a prototype-guided local update module to alleviate the client-drift issue, which introduces global prototypes as shared knowledge to force clients to learn consistent features. To reduce gradient conflicts between clients, we design a gradient refinement-based aggregation module to eliminate conflicting components between client gradients during global aggregation, thereby improving the agreement between clients. Experiments are conducted on a public dataset to verify FedAAR’s effectiveness, which consists of 87,621 2-s motion data. The results demonstrate that FedAAR outperforms state-of-the-arts, with precision (75.23%), recall (75.17%), F1-score (74.70%), and accuracy (88.88%), respectively. The ablation experiments also show FedAAR’s robustness against various factors (i.e., different data sizes, communication frequency, and client numbers).

Federated learning is a decentralized topology of deep learning, that trains a shared model through data distributed among each client (like mobile phones, wearable devices), in order to ensure data privacy by avoiding raw data exposed in data center (server). After each client computes a new model parameter by stochastic gradient decrease (SGD) based on their own local data, all locally-computed parameters will be aggregated in the server to generate an updated global model. Almost all current studies directly average different client computed parameters by default, but no one gives an explanation why averaging parameters is a good approach. In this paper, we treat each client computed parameter as a random vector because of the stochastic properties of SGD, and estimate mutual information between two client computed parameters at different training phases using two methods in two learning tasks. The results confirm the correlation between different clients and show an increasing trend of mutual information with training iteration. However, when we further compute the distance between client computed parameters, we find that parameters are getting more correlated while not getting closer. This phenomenon suggests that averaging parameters may not be the optimum way of aggregating trained parameters.

AI tools provide individualized learning experiences tailored to specific language learners' requirements. Artificial intelligence-driven programs can customize lessons, exercises, and feedback by evaluating a learner's progress and determining strengths and shortcomings. This individualized method aids students in concentrating on their areas of weakness, thus improving their total language competency. This narrative literature review aims to gather relevant literature discussing AI tools as supplementary support in aiding a learner's language acquisition. Relevant studies stated that there are a lot of chances to improve learning outcomes when AI tools are used in language acquisition. Teachers who want to give their students more resources and help should consider including these tools in their lessons. To preserve the interpersonal and cultural components of language learning, it is imperative to strike a balance between the application of AI and conventional teaching techniques. AI tools are always accessible, providing assistance to learners whenever needed. This is especially helpful for individuals with unpredictable schedules or restricted access to conventional language learning materials.

This paper explores the increasing use of big data in education to create personalized learning experiences tailored to individual student needs. The trend, driven by vast amounts of educational data, requires the development of new techniques to optimize learning outcomes while addressing the ethical implications of data collection and utilization. The paper emphasizes the necessity for robust policy frameworks to ensure transparency and trust, based on clear guidelines and parental consent. It also discusses how educational institutions are updating curricula to prepare learners for a rapidly evolving workforce, incorporating contemporary AI technologies that impact education systems. The dual influence of AI presents both opportunities and challenges, necessitating a balanced and thoughtful approach. The paper advocates for a holistic strategy in integrating AI in education, emphasizing the importance of collaboration between the AI industry and other sectors to develop comprehensive policy frameworks for the ethical and effective use of AI in education.

In this study, we present a federated learning approach that combines a multi-branch network and the Oort client selection algorithm to improve the performance of federated learning systems. This method successfully addresses the significant issue of non-iid data, a challenge not adequately tackled by the commonly used MFedAvg method. Additionally, one of the key innovations of this research is the introduction of uniformity, a metric that quantifies the disparity in training time amongst participants in a federated learning setup. This novel concept not only aids in identifying stragglers but also provides valuable insights into assessing the fairness and efficiency of the system. Experimental results underscore the merits of the integrated multi-branch network with the Oort client selection algorithm and highlight the crucial role of uniformity in designing and evaluating federated learning systems.

The NOSTR is a communication protocol for the social web, based on the w3c websockets standard. Although it is still in its infancy, it is well known as a social media protocol, thousands of trusted users and multiple user interfaces, offering a unique experience and enormous capabilities. To name a few, the NOSTR applications include but are not limited to direct messaging, file sharing, audio/video streaming, collaborative writing, blogging and data processing through distributed AI directories. In this work, we propose an approach that builds upon the existing protocol structure with end goal a decentralized marketplace for federated learning and LLM training. In this proposed design there are two parties: on one side there are customers who provide a dataset that they want to use for training an AI model. On the other side, there are service providers, who receive (parts of) the dataset, train the AI model, and for a payment as an exchange, they return the optimized AI model. The decentralized and censorship resistant features of the NOSTR enable the possibility of designing a fair and open marketplace for training AI models and LLMs.

Recently, federated learning (FL) has been receiving great attention as an effective machine learning method to avoid the security issue in raw data collection as well as to distribute the computing load to edge devices. However, even though wireless communications is a essential part for implementing FL in edge networks, there have been a few works that analyze the effect of wireless networks on FL. In this paper, we investigate FL in small-cell networks where multiple base stations (BSs) and users are located according to homogeneous Poisson point process (PPP) with different densities. We comprehensively analyze the effects of geographic node deployment on the model aggregation in FL on the basis of stochastic geometry-based analysis. We derive the closed-form expressions of coverage probability with tractable approximations and discuss the minimum required BS density for achieving a target model aggregation rate in small-cell networks. Our analysis and simulation results provide insightful information for understanding the behaviors of FL in small-cell networks, and it can be exploited as a guideline for designing the network facilitating the wireless FL.

Stress is a natural yet potentially harmful aspect of human life, necessitating effective management, particularly during overwhelming experiences. This paper presents a scoping review of personalized stress detection models using wearable technology. Employing the PRISMA-ScR framework for rigorous methodological structuring, we systematically analyzed literature from key databases including Scopus, IEEE Xplore, and PubMed. Our focus was on biosignals, AI methodologies, datasets, wearable devices, and real-world implementation challenges. The review presents an overview of stress and its biological mechanisms, details the methodology for the literature search, and synthesizes the findings. It shows that biosignals, especially EDA and PPG, are frequently utilized for stress detection and demonstrate potential reliability in multimodal settings. Evidence for a trend towards deep learning models was found, although the limited comparison with traditional methods calls for further research. Concerns arise regarding the representativeness of datasets and practical challenges in deploying wearable technologies, which include issues related to data quality and privacy. Future research should aim to develop comprehensive datasets and explore AI techniques that are not only accurate but also computationally efficient and user-centric, thereby closing the gap between theoretical models and practical applications to improve the effectiveness of stress detection systems in real scenarios.

This article aims to investigate the potential of artificial intelligence (AI) in revolutionizing the education sector and further advancing the vision of Education 6.0. This study explores how AI can foster personalized learning experiences and adaptive pedagogy. By analyzing recent developments and real-world applications of AI in education, this article offers insights into how AI can support educators in tailoring instruction to individual students' needs, promoting engagement, and optimizing learning outcomes. The article also discusses ethical considerations and challenges associated with the integration of AI in education, emphasizing the importance of responsible implementation and human supervision.

In the age of digital revolution technology is being used more and more in education around the world to improve student learning. By enabling ongoing skill development and knowledge acquisition through digital platforms e-learning has become a crucial tool. Through E-Learning online learning is becoming increasingly popular and effective in addition to traditional methods of instruction as highlighted by the latest statistics from Badan Pusat Statistik Indonesia (BPS). In order to address the issues of scalability, interactivity and security in e-learning platforms. This study introduces a web-based learning environment that is focused on teaching programming. Leveraging the MERN stack, while also integrating TypeScript for improved code reliability, Next.js for accelerated React.js functionality, Tailwind CSS for responsive design. Backend features include media management with Cloudinary, secure video streaming with VdoCipher, data storage with MongoDB, and caching with Redis. Iterative development will be following agile methodology. In order to create a vibrant and encouraging learning environment the platform includes interactive course pages multimedia content real-time feedback mechanisms and community-building tools. Despite some early difficulties the implementation phase effectively achieved development goals. To further support international online learning initiatives further efforts will be directed toward improving platform security and functionality.

Federated unlearning represents a sophisticated evolution in the domain of machine learning, particularly within federated learning frameworks. In financial applications, where data privacy and security are paramount, federated unlearning allows institutions to selectively remove or “unlearn” specific data from trained models without compromising their overall performance and accuracy. This capability is essential for ensuring that models remain adaptable, secure, and compliant with regulatory requirements, while minimizing the need for expensive retraining. In this paper, we explore various financial applications where federated unlearning can have a significant impact, including fraud detection, portfolio management, and credit risk modeling. By allowing targeted removal of outdated, erroneous, or sensitive data, federated unlearning enhances the agility of financial models, enabling institutions to keep pace with the dynamic financial landscape. Using practical numerical examples, we demonstrate how unlearning improves model accuracy and decision-making while maintaining data privacy across distributed systems. This paper underscores the critical role of federated unlearning in addressing the challenges of modern financial institutions, offering insights into its practical applications and future potential.

In recent years, the scientific and technological developments led to an explosion of available videos on the web, increasing the necessity of fast and effective video analysis and summarization. Video summarization methods aim to generate a synopsis by selecting the most informative parts of the video content. The user’s personal preferences, often involved in the expected results, should be taken into account in the video summaries. In this paper, we provide the first comprehensive survey on personalized video summarization relevant to the techniques and datasets used. In this context, we classify and review personalized video summary techniques based on the type of personalized summary, on criteria, on the video domain, on the source of information, on the time of summarization, and on the machine learning technique. Depending on the type of methodology used by the personalized video summarization techniques for the summary production process, we classify the techniques into five major categories, which are feature-based video summarization, key frame selection, shot selection-based approach, video summarization using trajectory analysis, and personalized video summarization using clustering. We also compare personalized video summarization methods and present 37 datasets used to evaluate personalized video summarization methods. Finally, we analyze opportunities and challenges in the field and suggest innovative research lines.

The genetic causes of the differentiated, highly treatable and mostly non-fatal papillary thyroid cancer (PTC) are not yet fully known. The widely accepted PTC etiology blames altered sequence or/and expression level of certain biomarker genes. However, our previous studies found that the PTC biomarkers played significantly lower roles than the personalized PTC Gene Master Regulators. Our publicly accessible gene expression data from the cancer nodule and the surrounding normal tissue of surgically removed PTC tumor were re-analyzed to determine the transcriptomic effects on the genomic fabrics responsible for the DNA replication, repair and transcription. Tumor results were compared to the gene expression profiles of the papillary (BCPAP) and the anaplastic (8505C) human thyroid cancer cell lines. The analyses were carried out from the Genomic Fabric Paradigm (GFP) perspective that provides the most theoretically possible comprehensive characterization of the transcriptome and its alterations in disease. The study indicated that, in addition to regulating numerous genes, changes in the homeostatic control of transcript abundances and remodeling of the gene networks had major contributions to the PTC occurrence, resilience and proliferation. Therefore, the molecular mechanisms responsible for gene expression control and inter-coordination should also be considered when designing personalized anti-cancer gene therapies.

This paper examines the learning experience of an inter-faculty higher education course conducted in an expanded remote format, using digital space as facilitator. The study focuses on the implementation of personalized education, a pedagogical approach that emphasizes the diversification of resources, strategies, and actions to empower learners in assigning personal value and meaning to their learning journey. By presenting an action research project, this study highlights the pivotal role of personalized education in guiding the course design. To foster an engaging learning environment, the course incorporated open activities and leveraged tools and platforms of the social web. Through these means, the participants, forming a community of practice, were able to collaboratively generate knowledge before, during, and after synchronous sessions. Furthermore, the participants effectively disseminated this knowledge through networks and engaged in discussions with external stakeholders. The study underscores how personalized education facilitates situated and meaningful learning by fostering expanded, flexible, and distributed experiences. These experiences revolve around a transmedia narrative that seamlessly integrates digital culture into learning opportunities and processes, effectively naturalizing its presence. The results highlight the value of personalized education approaches in adapting to the complexities of the digital age and establishing meaningful learning experiences for students.

Recent advances in organoid technology have revolutionized cancer biology and thera-peutic interventions, offering personalized immunotherapy treatment. Organoids, three-dimensional cell cultures derived from patient tumors, accurately replicate the tumor microenvi-ronment, providing unprecedented insights into tumor-immune interactions and therapeutic re-sponses. In this literature-based study, we discuss various culture methods for the diverse appli-cations of organoids in cancer immunotherapy, including drug screening, personalized treatment strategies, and mechanistic studies. Additionally, we address the technological challenges asso-ciated with these methods and propose potential future solutions to accelerate the development of novel immunotherapeutic approaches. This review highlights the transformative potential of or-ganoid models in advancing preclinical cancer immunotherapy modeling, screening, and evalua-tion, paving the way for more effective and personalized cancer treatments.

This paper presents an on-device semi-supervised human activity detection system that can learn and predict human activity patterns in real time. The proposed semi-supervised learning (SSL) framework uses sparsely labelled user activity events acquired from Inertial Measurement Unit (IMU) sensors installed as wearable device. The objective is to learn classification of multiple classes of human activities in real-time. The proposed cluster-based learning model in this approach is trained with data from the same target user, thus preserving data privacy while providing personalized activity detection services. Two different cluster labelling strategies, namely, population-based, and distance-based, are employed to achieve the desired classification performance. A comparative study of these strategies has been presented in terms of classifiability and classification accuracies. The proposed system is shown to be computationally efficient, which is relevant in the context of limited computing resources on typical wearable devices. The trade-off between classification accuracy and computation complexity is analyzed for different algorithmic hyper-parameters and other system parameters. Extensive experimentation and simulation study have been conducted on a multi-user human activity data from the public do-main in order to validate the proposed learning paradigm.

A framework for analysis that focusses on options for structuring and funding the State Owned Enterprises (SOEs) is required in order to apply concepts and information more effectively to policy analysis and program creation. Regardless of the political or economic driving reason for SOE organisation, the requirements of the people must be met in developing countries necessitate that the strategy be viewed from this angle. The types of SOEs and the applicability of fiscal federalism are examined in this article. There are certain lessons to be learnt from the past, and a bad policy can have unintended repercussions. Strategies must be developed for the organisation of SOEs while taking into account the various starting situations.

Federated Distillation (FD) has recently attracted increasing attention for its efficiency in aggregating multiple diverse local models trained from statistically heterogeneous data of distributed clients. Existing FD methods generally treat these models equally by merely computing the average of their output soft predictions for some given input distillation sample, which does not take the diversity across all local models into account, thus leading to degraded performance of the aggregated model, especially when some local models learn little knowledge about the sample. In this paper, we propose a new perspective that treats the local data in each client as a specific domain and design a novel domain knowledge aware federated distillation method, dubbed DaFKD, that can discern the importance of each model to the distillation sample, and thus is able to optimize the ensemble of soft predictions from diverse models. Specifically, we employ a domain discriminator for each client, which is trained to identify the correlation factor between the sample and the corresponding domain. Then, to facilitate the training of the domain discriminator while saving communication costs, we propose sharing its partial parameters with the classification model. Extensive experiments on various datasets and settings show that the proposed method can improve the model accuracy by up to $6.02\%$ compared to state-of-the-art baselines.

This paper presents an innovative methodology for enhancing museum visitor experiences through personalized content delivery using a combination of explicit and implicit visitor profiling. The approach integrates Bluetooth Low Energy (BLE) smart badges to collect telemetry data, enabling precise visitor localization and dynamic group formation based on real-time proximity and shared interests. Initial profiling begins with OAuth registration and brief surveys, which is then refined through continuous tracking of exhibit interactions and time spent at each exhibit. An AI-driven system delivers content to individual and group profiles, fostering both personalized learning and social interaction. This methodology addresses the limitations of traditional profiling by adapting to visitor behaviors in real-time while maintaining a strong focus on data privacy and ethical considerations. The proposed system not only enhances engagement and satisfaction but also sets the stage for future advancements in personalized cultural experiences.

In the current era of emerging technologies and cutting-edge inventions, it is equally important to consider the ethical aspects of research and innovations to have an ethical space for the sustainable progress and safety of humanity. The field of nanotechnology research and innovation often faces criticism due to its potential adverse impacts on health, the environment, limited accessibility, high costs, and concerns about data privacy. Therefore, it is crucial to address and resolve these ethical issues systematically. In this regard, we present our findings and methods for exploring the application of the ethics technology assessment method by considering its relevance at different stages of innovation development. Specifically, the present work provides a conceptual framework for ethics technology assessment focused on evaluating a significant innovation in nanotechnology within the domain of food safety with specific reference to mycotoxin remediation. By leveraging the value-sensitive design approach, we described various strategies for analyzing the ethical implications of nanotechnology-based innovations in mycotoxin remediation. Through comprehensive ethics technology assessment, nanotechnology-based innovations can be guided toward becoming more value-oriented, thereby making substantial contributions to social progress.

Manual material handling (MMH) significantly impacts worker health and productivity, often leading to musculoskeletal disorders (MSDs) primarily in the lower back. As a novel assistive technology, exoskeletons may serve as ergonomic tools to mitigate these work-related MSDs. It is essential to examine exoskeletons from the users’ perspectives before their widespread implementation in occupational settings. This study investigates the effectiveness of a passive back-support exoskeleton (BExo) in reducing perceived physical exertion and improving ergonomic safety in manufacturing context. Twenty-two college students were recruited to perform MMH tasks in a controlled lab environment, both with and without the BExo, followed by completing a survey questionnaire on various aspects of the BExo. Using ANOVA, the study analyzed biomechanical exertion across various body parts and tasks. The findings indicate that the BExo substantially alleviates discomfort in the low back, shoulders and knees, thereby, enhancing ergonomic posture and reducing fatigue. These results underscore the potential of passive exoskeletons to boost workers’ safety and efficiency, providing valuable insights for future ergonomic strategies in industrial setting.

Analytical data platform have been used for decades for improving the organizational performance. Starting from the data warehouses used primarily for structured data processing, through the data lakes oriented for raw data storage and post-hoc data analyses, to the data lakehouses − a combination of raw storage and business intelligence pre-processing for improving the platform efficacy. But in recent years a new architecture called Data Mesh has emerged. The main promise of this architecture is to remove the barriers between operational and analytical teams in order to boost the overall value extraction from the big data. A number of attempts has been made to formalise and to implement it in existing projects. Although being defined as a socio-technical paradigm, data mesh still lacks the technology support for enabling its widespread adoption. To overcome this limitation, we propose a new view of the platform requirements alongside the formal governance definition that we believe can help in successful adoption of the data mesh. It is based on fundamental aspects such as decentralized data domains and federated computational governance. In addition, we also present a blockchain-based implementation of a mesh platform as a practical validation to our theoretical proposal. Overall, this article demonstrates a novel research direction of the information system decentralization technologies.

In the design of real-world networks, researchers evaluate various structural parameters to assess vulnerability, including connectivity, toughness, and tenacity. Recently, the tightness metric has emerged as a potentially superior vulnerability measure, although many related theorems remain unknown due to its novelty. Harary graphs, known for their maximum connectivity, are an important class of graph models for network design. Prior work has evaluated the vulnerability of three types of Harary graphs using different parameters, but the tightness metric has not been thoroughly explored. This article aims to calculate the tightness values for all three types of Harary graphs. First, it will attempt to calculate the lower bound for the value of the tightness parameter in Harary graphs using existing lemmas and theorems. Then, by presenting new lemmas and theorems, we will try to find the exact value or upper bound for this parameter in Harary graphs. For the first type of Harary graph, the tightness is precisely determined, while for the second and third types, upper bounds are provided due to structural complexity. The lemmas, theorems, and proof methods presented in this research may be used to calculate other graph and network parameters. However, the newness of the tightness parameter means that further research is needed to fully characterize its properties.

Dental agenesis is one of the most common developmental anomaly in humans and it is frequently as-sociated with several other oral abnormalities. The present case describes nonfamilial agenesis of per-manent teeth in a twenty-one-year-old boy, with no apparent systemic abnormalities. The treatment in-cluded a personalized and interdisciplinary approach by the means of endodontics, orthodontics, im-plants supported restorations and prosthetic treatments. The treatment plan was thoroughly elaborated using photographic analysis, study models, orthopantomogram, CBCT and cephalograms. Virtual smile design, diagnostic waxing, and mock-up previsualized the treatment objectives. The edentulous spaces were reconstructed by inserting dental implants, and monolithic Zirconia implant supported restorations. The final results showed a highly aesthetic and functional rehabilitation. Periodic checkups have shown that the stability of the result is well maintained and that the implant supported restorations are an op-timal solution for patients with multiple anodontia.

Lung cancer (LC) continues to be an important public health problem being the most common form of cancer and a major cause of cancer deaths worldwide. Despite the great bulk of research to identify genetic susceptibility genes by genome-wide association studies, only few loci associated to nicotine dependence have been consistently replicated. Our previously published study in few phenotypically discordant sib-pairs identified a combination of germline truncating mutations in known cancer susceptibility genes in never smoker early-onset LC patient does not present in their healthy sib. These results firstly demonstrated the presence of an oligogenic combination of disrupted cancer predisposing genes in non-smokers patients, giving experimental support to a model of a “private genetic epidemiology”. Here, we used a combination of whole-exome and RNA sequencing coupled with a discordant sib’s model in a novel cohort of pairs of never-smokers early-onset LC patients and in their healthy sibs used as controls. We selected rare germline variants predicted as deleterious by CADD and SVM bio-informatics tools and absent in the healthy sib. Overall, we identified an average of 200 variants per patient, about 10 of which in cancer predisposing genes. In most of them, RNA sequencing data reinforced the pathogenic role of the identified variants showing: i) downregulation in LC tissue (indicating “second hit” in tumor suppressor genes); ii) upregulation in cancer tissue (likely oncogene); iii) downregulation in both normal and cancer tissue (indicating transcript instability). The combination of the two techniques demonstrates that each patient has an average of six (with a range from four to eight) private mutations with functional effect in tumor predisposing genes. The presence of a unique combination of disrupting events in the affected subjects may explain the absence of familial clustering of non-small cell lung cancer. In conclusion, these findings indicate that each patient has his/her own “predisposing signature” to cancer development and suggest the use of personalized therapeutic strategies in lung cancer.

In federated learning (FL), clients train models locally without sharing raw data, ensuring data privacy. In particular, federated distillation transfers knowledge to clients regardless of the model architecture. However, when groups of clients with different data distributions exist, sharing the same knowledge among all clients becomes impractical. To address this issue, this paper presents an approach that clusters clients based on the output of a client model trained using their own data. The clients are clustered based on the predictions of their models for each label on a public dataset. Prior knowledge of the number of groups is not required. Evaluations on the MNIST dataset showed that our method accurately identified these group structures and improved accuracy by 15–75% compared with traditional federated distillation algorithms when distinct group structures were present. Additionally, we observed significant performance improvements in smaller client groups, bringing us closer to fair FL.

This paper discusses the application of artificial intelligence in imaging omics, especially in cancer research. Imaging omics enables detailed analysis of spatial and temporal heterogeneity of tumours through high-throughput extraction of quantitative features from medical images such as MRI, PET, and CT. This paper focuses on applying PARKS systems to automate the recognition, segmentation, and extraction of image features, significantly enhancing the capabilities of clinical decision support systems (CDSS). The future direction is to establish a robust network infrastructure for radiology Medication-led Health care (RLHC) to facilitate the development and application of personalised treatment protocols, and to improve diagnostic accuracy, prognosis assessment, and treatment recommendations by uploading quantitative image features to a shared database and comparing them with historical images.

The field of personalized medicine is undergoing a transformative shift through the integration of multi-omics data, which encompasses genomics, transcriptomics, proteomics, and metabolomics. This synergy allows for a comprehensive understanding of individual health by analyzing genetic, molecular, and biochemical profiles. The generation and integration of multi-omics data enable more precise and tailored therapeutic strategies, improving the efficacy of treatments and reducing adverse effects. However, several challenges hinder the full realization of personalized medicine. Key hurdles include the complexity of data integration across different omics layers, the need for advanced computational tools, and the high cost of comprehensive data generation. Additionally, issues related to data privacy, standardization, and the need for robust validation in diverse populations remain significant obstacles. Looking ahead, the future of personalized medicine promises advancements in technology and methodologies that will address these challenges. Emerging innovations in data analytics, machine learning, and high-throughput sequencing are expected to enhance the integration of multi-omics data, making personalized medicine more accessible and effective. Collaborative efforts among researchers, clinicians, and industry stakeholders are crucial to overcoming these hurdles and fully harnessing the potential of multi-omics for individualized healthcare.

Human activity recognition (HAR) using wearable sensors is gaining popularity in the healthcare domain, enabling continuous monitoring and timely interventions. The conventional HAR systems are centralised and need data sharing with a server for processing and model training. However, this data sharing leads to various challenges, such as privacy concerns, increased communication and storage costs, inefficient computational algorithms, data scarcity due to unwillingness to share, and difficulty achieving high accuracy and personalised models simultaneously. To overcome these challenges, this paper introduces a privacy and energy-aware federated learning (FL) framework for model training without data sharing. The proposed scheme incorporates spiking neural networks (SNNs), which are particularly suited for resource-constrained devices due to their event-driven information processing. The key idea is to synergise the strengths of SNNs and long short-term memory (LSTM) to create a hybrid model for robust HAR. The proposed hybrid spiking LSTM (S-LSTM) model uses the LSTM as an input layer to capture temporal dependencies efficiently, followed by spiking layers that replace the typical activation using leaky integrate-and-fire neurons. We employ the combination of surrogate gradient learning and backpropagation through time (BPTT) to update the weights of SNN layers in a supervised manner. The performance of the proposed model is evaluated using two publicly available datasets for both indoor and outdoor scenarios. The results show that the proposed S-LSTM performs much better than spike convolutional neural networks (S-CNN) and simple CNN, achieving the accuracy of 97% and 87% for indoor and outdoor datasets, respectively. Personalisation, also known as fine-tuning the global model using local data, is crucial in HAR. It allows the model to adapt to individual behaviour and preferences, making predictions more accurate. The results also show that personalisation improves the performance of participants up to 9% on average. Additionally, analysis with 50% random client participation reduces communication costs by 50% with minimal impact on accuracy.

The shift towards Personalized Cancer Medicine (PCM) represents a significant transformation in cancer care, emphasizing tailored treatments based on genetic understanding of cancer at the cellular level. This review draws on recent literature to explore key factors influencing PCM implementation, highlighting the role of innovative leadership, interdisciplinary collaboration, and coordinated funding and regulatory strategies. Success in PCM relies on overcoming challenges such as integrating diverse medical disciplines, securing sustainable investment for shared infrastructures, and navigating complex regulatory landscapes. Effective leadership is crucial for fostering a culture of innovation and teamwork, essential for translating complex biological insights into personalized treatment strategies. The transition to PCM necessitates not only organizational adaptation but also the development of new professional roles and training programs, underscoring the need for a multidisciplinary approach and the importance of team science in overcoming the limitations of traditional medical paradigms. The conclusion underscores that PCM's success hinges on creating collaborative environments that support innovation, adaptability, and shared vision among all stakeholders involved in cancer care.

Background: Assessing the impact of bone health clinical management in breast cancer (BC) patients receiving adjuvant endocrine therapy and designing a personalized clinical pathway to reduce bone loss, in an Italian research hospital. Methods: The primary endpoint was to assess (through the process improvement organizational method) the clinical pathway that post-surgical BC patients prescribed with endocrine therapy undergo to prevent bone loss. The secondary endpoint was to design a personalized clinical pathway for a prompt implementation of guidelines, to assess and possibly prescribe antiresorptive therapy. Results: During the first year of the new diagnostic therapeutic assistance pathway execution, an 60% increase in Dual-Energy X-ray Absorptiometry evaluations within 30 days and a 39.5% increase in antiresorptive therapy prescription within 90 days (since prescription of endocrine therapy) were shown, thus increasing patients’ compliance. Conclusion: Case managers and bone health specialists in this context can improve patients’ adherence to therapies and bone health, helping physicians to expand their collaboration.

Rectal cancer presents a significant burden globally, often requiring multimodal therapy for locally advanced cases. Long-course chemoradiotherapy (LCRT) and short-course radiotherapy (SCRT) followed by surgery have been conventional neoadjuvant approaches. Recent trials favor LCRT due to improved local control. However, tumor distant recurrence remains a concern, prompting exploration of Total Neoadjuvant Therapy (TNT) as a comprehensive treatment strat-egy. Immune checkpoint inhibitors (ICIs) show promise, particularly in mismatch repair-deficient (dMMR) or microsatellite instability-high (MSI-H) tumors, potentially revolutionizing neoadjuvant regimens. Non-operative management (NOM) represents a viable alternative post-neoadjuvant therapy for selected patients achieving clinical complete response (cCR). Additionally, monitoring minimal residual disease (MRD) using circulating tumor DNA (ctDNA) emerges as a non-invasive method for treatment response assessment. This review synthesizes current evidence on TNT, ICIs, NOM and ctDNA, elucidating their implications for rectal cancer management and highlighting avenues for future research and clinical application.

It is estimated that pancreatic cancer will be the 2^nd leading cause of cancer-related deaths globally by 2030, highlighting the ongoing lack of effective treatment options in this devastating condition. There is a lack of reliable prognostic or predictive markers in pancreatic cancer to guide management decisions, whether for systemic chemotherapy, molecularly targeted therapies, or immunotherapies. To date, the results for targeted agents and immunotherapies in unselected populations of chemo-refractory pancreatic cancer have not met expectations. The reasons for this lack of efficacy of immunotherapy in pancreatic cancer are incompletely understood. The challenges in pancreatic cancer include the physical barrier created by the dense desmoplastic stroma surrounding the tumor, chemokine-mediated exclusion of T cells, poor antigenicity, paucity of infiltrating T cells within the tumor, ultimately leading to an immunosuppressive microenvironment. A better understanding of the role of inflammation in pancreatic cancer, its tumor microenvironment and individualized patient-related features, be they molecular, clinical or histopathological would enable a more effective tailored approach to the management of pancreatic cancer. In this review, the role of inflammation, the immune tumor microenvironment and potential immune biomarkers in pancreatic cancer are explored.

Despite widespread vaccination, early treatments, and improved understanding of the disease, the effects of SARS-CoV-2 infection remain significant worldwide. Many patients still suffer from severe COVID-19, necessitating admission to intensive care units. Remdesivir is a primary treatment option among viral RNA polymerase inhibitors for hospitalized SARS-CoV-2 patients. However, there is a lack of studies examining factors influencing its effectiveness in this context. We conducted a retrospective study throughout 2022, analyzing clinical, laboratory, and sociodemographic data from 252 hospitalized COVID-19 patients treated with remdesivir. Six machine learning algorithms were compared and validated to predict factors associated with the loss of clinical benefit from remdesivir in terms of mortality and hospital stay. Data were extracted from electronic health records. The eXtreme Gradient Boost (XGB) method achieved the highest balanced accuracy for both mortality (95.45%) and hospital stay (94.24%). Factors associated with worse outcomes of remdesivir use in terms of mortality included limitation of life support treatment, need for ventilatory support (especially invasive mechanical ventilation) on day 14 after the first dose of remdesivir, lymphopenia, low levels of albumin and hemoglobin, presence of flu and/or coinfection, and cough. Factors associated with worse outcomes of remdesivir use in terms of hospital stay included the number of doses of the COVID-19 vaccine, patchy lung density, bilateral pulmonary radiological status, number of comorbidities, oxygen therapy, troponin and lactate dehydrogenase levels, and asthenia. These findings highlight the effectiveness of XGB as a strong candidate for accurately categorizing COVID-19 patients undergoing remdesivir treatment.

Ethnopharmacology experts face several challenges when identifying and retrieving documents and resources related to their scientific focus. The volume of sources that need to be monitored, the variety of formats utilized, the different quality of language use across sources, present some of what we call “big data” challenges in the analysis of this data. This study aims to understand if and how experts can be supported effectively through intelligent tools in the task of ethnopharmacological literature research. To this end, we utilize a real case study of ethnopharmacology research, aimed at the Southern Balkans and Coastal zone of Asia Minor. Thus, we propose a methodology for more efficient research in ethnopharmacology. Our work follows an “Expert-Apprentice” paradigm in an automatic URL extraction process, through crawling, where the apprentice is a Machine Learning (ML) algorithm, utilizing a combination of Active Learning (AL) and Reinforcement Learning (RL), and the Expert is the human researcher. ML-powered research improved 3.1 times the effectiveness and 5.14 times the efficiency of the domain expert, fetching a total number of 420 relevant ethnopharmacological documents in only 7 hours versus an estimated 36-hour human-expert effort. Therefore, utilizing Artificial Intelligence (AI) tools to support the researcher can boost the efficiency and effectiveness of the identification and retrieval of appropriate documents.

This study addresses the challenge of treating skin cancer effectively, especially in visible areas like the face, where traditional methods may be undesirable. Current options, including surgery and radiation therapy, have limitations such as complexity, expense, and imprecise dose delivery. The proposed solution involves additively manufacturing personalized patches using diagnostic imaging data, offering a targeted and controlled dose of beta-emitting isotopes like Y-90. The patches, tailored to the tumor's shape and depth, aim to minimize damage to healthy tissues and structures. Through 3D printing technology we demonstrated the feasibility of creating individualized applicators filled with radioisotopes, that have the potential to deliver therapeutic doses while ensuring patient comfort and affordability. This innovative approach, based on radiation delivery simulations, offers a promising future alternative to conventional radiation therapy methods. Clinical trials are required to validate its efficacy and safety, but the potential for reduced radiation exposure and enhanced treatment precision positions this personalized and targeted therapy as a valuable advancement in skin cancer treatment.

Psoriasis is a chronic, multisystem, inflammatory disease. The typical clinical cutaneous manifestation of psoriasis is scaly erythema or plaque, limited or widely distributed. However, psoriasis is far beyond the skin involvement and provides many challenges, including associated comorbidities. In this review, we classify psoriasis as cutaneous and systemic psoriasis, according to the clinical diversity and associated comorbid diseases, and recommend classification criteria for psoriasis. The key objective of this novel classification is to raise awareness of the complexity of this multifaceted disease and help to better understand and manage this complex disease comprehensively.

Twenty-first century infrastructure needs to respond to changing demographics, becoming climate neutral, resilient, and economically affordable, while remaining a driver for development and shared prosperity. However, the infrastructure sector remains one of the least innovative and digitalized, plagued by delays, cost overruns, and benefit shortfalls [1-4]. The root cause is the prevailing fragmentation of the infrastructure value chain [5]. To support overcoming the shortcomings, an integration of the value chain is needed. This could be achieved through a use-cased-based creation of federated digital platforms applied to infrastructure projects. Such digital platforms enable full-lifecycle participation and responsible governance guided by a shared infrastructure vision.

Sweat rate and electrolyte losses have a large inter-individual variability, a personalized approach to hydration can overcome this issue to meet an individual’s needs. This study aimed to investigate the effects of personalized hydration strategy (PHS) on fluid balance and intermittent exercise performance. Twelve participants conducted 11 laboratory visits including a VO2max test and two 5-day trial arms under normothermic (NOR) or hyperthermic (HYP) environmental conditions. Each arm began with 3-day of familiarization exercise following by two random exercise trials with a either PHS or control (CON). Then, participants crossed-over to the second arm for: NOR+PHS, NOR+CON, HYP+PHS, and HYP+CON. The PHS was prescribed according to participants’ fluid and sweat sodium losses. CON drank ad-libitum of commercially-available electrolyte solution. Exercise trials consisted of two phases: 1) 45min constant workload; 2) high-intensity intermittent exercise (HIIT) until exhaustion. PHS had a significantly greater fluid intake compared to CON. HYP+CON produced the lowest sweat sodium concentration. HYP+PHS had a lower thirst perception and a longer HIIT compared to HYP+CON. Thus, PHS reinforces fluid intake and successfully optimizes hydration status, regardless of environmental conditions. PHS may be or is an important factor to preventing negative physiological consequences during high-intensity exercise in the heat.

There is currently no effective treatment strategy for the recurrent/metastatic adenoid cystic carcinoma (R/M ACC). Furthermore, recent single-agent and combination immunotherapy trials have failed in unselected ACC cohorts unlike non-ACC salivary gland cancers. Genomic profiling revealed no actionable targets but NOTCH1 and KDM6A frameshift and CTCF splice site mutations (no MYB/L fusion) with a low TMB, MSS and negative PD-L1. We recommended anti-PD-1 plus anti-CTLA-4 combination based on TMB 2-fold greater than median TMB in ACC, tumor harboring multiple immunogenic frameshift or splice site mutations, and PD-L1 negativity. Accordingly, we achieved a complete response in a radiotherapy and chemotherapy-refractory patient with locally recurrent lacrimal gland ACC and lung metastasis following personalized immunotherapy in combination with integrative therapeutics. Therefore, it is crucial to assess not only conventional immune biomarkers but also patient-specific parameters.

Background/Objectives: The global burden of colorectal cancer (CRC) is projected to increase by 2040, highlighting the need for prevention strategies, early detection and adequate follow. The aim of this study was to provide an overview of the progress in personalized medicine of CRC in Serbia, with results and insights from the Institute for Oncology and Radiology of Serbia (IORS), and to propose guidance for tackling observed challenges in the future. Methods: Epidemiological data was derived from official global and national cancer registries and IORS electronic medical records. Germline genetic testing for Lynch syndrome was performed by Next Generation Sequencing. RAS and BRAF mutation analyses were performed using qPCR diagnostic kits. Results: Epidemiology and risk factors, prevention and early detection programs, as well as treatment options and scientific advances have been described in detail. Out of 103 patients who underwent germline testing for Lynch syndrome, 19 (18.4%) had a pathogenic or likely pathogenic mutation in one of the MMR genes, and 8 (7.8%) in other CRC-associated genes (APC, CHEK2, MUTYH). Of 6369 tested patients, 50.43% had a mutation in KRAS or NRAS genes, while 9.54% had the V600 mutation in the BRAF gene. Conclusions: Although significant improvements in CRC management have occurred globally in recent years, the high incidence of young-onset CRC and the growing elderly population due to a rise in life expectancy require a strategic approach leading to population-based systemic solutions, especially important for countries with limited health care resources as Serbia.

Today, LLMs are not good at personalization providing recommendation. They advise physicians and financial advisors to ask professionals in respective fields for help, even having user information available. Answering questions of software professionals, LLM needs to deliver in-depth answers with codes or algorithms, whereas for professionals in other fields would need definitions and main concepts. The intent of this chapter is to make LLM answer tailored to the needs of users, taking into account available information about them. To do that, we need to generalize available information about a person like her health record, maintaining the privacy of this person. We rely on meta-learning techniques to design a LLM prompt to produce a personalization prompt to obtain a suitable relevant information. Such “meta-prompt” is produced by generalization operation applied to available documents for the user. These documents need to be de-identified so that they are sufficient for personalization on one hand and will maintain user privacy on the other hand. The second neuro-symbolic technique to support personalization is abductive reasoning, acting in parallel to LLM fine-tuning. Traditional recommendation and personalization techniques as well as modern, deep learning – based are presented, and the comparison is drawn to the proposed approach. We also share the evaluation and comparative analyses of these approaches. We consider an example for how to build personalization LLM systems coming from Langchain platform. We will explore how to construct chains to form a personalization profile for a user and apply it to user search and recommendation requests.

Background: to assess the benefit of a deep inspiration breath hold (DIBH) over standard irradiation technique, and eventually to identify anatomical and/or treatment preplanning characteristics correlated with LAD dose. Methods: We retrospectively identified patients with left-sided breast cancer who underwent whole breast radiotherapy in DIBH. All patients had both plans in DIBH and free-breathing (FB). Receiving operating characteristics (ROC analysis) were performed to define the cut-off point of parameters to predict LAD maximum dose >10 Gy and LAD mean dose > 4 Gy. Areas under the curve (AUCs) were calculated for all variables. Post-test probability has been calculated to evaluate advantage for parameters combination. Results: One hundred ninety-seven patients were identified. LAD dose was significantly reduced in DIBH plans with maximum and mean dose reduced by 31.7% (mean value 3.5 Gy vs 4.8 Gy, p=<0.001) and 28.1% (mean value 8.2 Gy vs 12.8 Gy, p=<0.001) in DIBH plans compared to FB plans. The strongest predictor of LAD maximum dose > 10 Gy and a LAD mean dose > 4 Gy was the minimum distance of LAD from tangent open fields. Other parameters were lung volume and heart volume for LAD Dmax > 10 Gy and lung volume, heart volume and breast separation for LAD Dmean > 4 Gy. Conclusion: The dosimetric benefit of DIBH is valid for all patients and DIBH should be preferred for all left sided patients.

Background: Immunologic and neuro-inflammatory pathways have been found to play a major role in the pathogenesis of many neurological disorders such as epilepsy, proposing the use of novel therapeutic strategies. In the era of personalized medicine and in the face of the exhaustion of anti-seizure therapeutic resources it is worth looking at the current or future possibilities that neuroimmunomodulator or anti-inflammatory therapy can offer us in the management of patients with epilepsy. Methods: We performed a narrative review on the recent advances on the basic epileptogenic mechanisms related to the activation of immunity or neuroinflammation with special attention to current and future opportunities for novel treatments in epilepsy. Results: Neuroinflammation can be considered a universal phenomenon and occurs in structural, infectious, post-traumatic, autoimmune, or even genetically based epilepsies. The emerging research developed in recent years has allowed us to identify the main molecular pathways in-volved in these processes. These molecular pathways could constitute future therapeutic targets for epilepsy. Conclusions: Different drugs current or in development have demonstrated their capacity to inhibit or modulate molecular pathways involved in the immunologic or neuroinflammatory mechanisms described in epilepsy. Some of them should be tested in the future as possible antiepileptic drugs.

Gestational diabetes mellitus (GDM) triggers alterations in the maternal gut microbiome. Alongside metabolic shifts, microbial products may impact clinical factors and influence pregnancy outcomes. Here, we performed two separate analyses to integrate the maternal microbiome-metabolomic differences from a subset of the “Choosing Healthy Options in Carbohydrate Energy” (CHOICE) study. Women diagnosed with GDM were randomized to a diet higher in complex carbohydrates (CHOICE, n=18, 60% complex carbohydrate/25% fat/15% protein) or a conventional GDM diet (CONV, n=16, 40% carbohydrate/45% fat/15% protein). All meals were provided, diets were eucaloric, and fiber content was similar. CHOICE diet was associated with increases in TMAO, lysophosphotidylcholines, indoxyl sulfate, and several triglyceride (TG) species, while CONV diet was associated with higher plasma hippuric acid, aspartic acid, betaine, glutamic acid, indole propionic acid, and indoleacetic acid, suggesting that differences in lipid metabolism and tryptophan utilization pathways are linked to specific maternal diets. We also identified two latent metabolic groups not associated with either diet treatment: one metabolome associated with higher fasting glucose, and the other associated with higher fasting TGs after 7 weeks on either diet. Our study underscores the importance of utilizing dietary approaches to optimize maternal-fetal health outcomes in GDM treatment.

Background: The hub-and-spoke model (HSM), offers a framework for efficient healthcare service delivery. This scoping review seeks to establish the implementation and effectiveness of the HSM in Antimicrobial Stewardship (AMS) programs. Methods: The Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR) reporting guideline was followed. A systematic search was conducted in four electronic databases (PubMed, Medline, Cochrane library and Google scholar) from inception until January 2024. Studies where the HSM was used for delivering any AMS activity, intervention or action, were included. No study specific filters were applied and all populations, study designs and health settings were included. Data screening and selection was achieved using Rayyan. Three authors independently screened studies, with conflicts resolved by a fourth author. Data were narratively synthesized. Standard appraisal tools were impractical; however, critically evaluation of data collection and outcome reporting was ensured. Results: Out of 1,438 articles, three were included in the scoping review. The primary interventions utilising the HSM in AMS involved reducing antibiotic misuse, training healthcare professionals, case-based learning, establishing AMS programs, developing antibiograms, and formulating policies or guidelines pertinent to AMS. The studies demonstrated significant clinical improvements in AMS. Clinical outcomes from the studies include a significant reduction in antimicrobial usage and improved antibiotic management, with a notable decrease in days on antimicrobial therapy and increased antibiotic de-escalation. Key facilitators for AMS program success: tailored education, collaborative learning, strong leadership, strategic practices, and data-driven decisions. Key barriers: leadership challenges, change resistance, knowledge gaps, inadequate data systems, resource limitations, and technological constraints. Conclusion: The review identified a literature gap in HSM use for AMS programs. Further studies are needed to assess HSM's effectiveness, feasibility, and cost-effectiveness in AMS contexts.

Considering the multifaceted nature of neurodegenerative diseases like dementia and the neces-sity for interprofessional knowledge, this research extends its scope to encompass professionals with diverse levels of training and experience in dementia care. A need analysis for the project “My INdividual Digital EDucation.RUHR” (MINDED.RUHR) is conducted to develop an au-tomatized recommender system for individual learning content using AI. In this sub-study, the aim was to reveal didactic specialties, knowledge gaps and structural challenges of further edu-cation in dementia care of different health professions and to derive learning preference personae. Eight focus group interviews among nine health professions and up to six participants (n = 34) each took place to survey distinct didactic experiences and learning needs. The results reflect various learning preferences, with a propensity to multimedia, practical and interactive tasks. Health professions are used to digital education but show aversions against certain e-learning formats. The derived learning preference personae constitute profound blueprints for a us-er-centred digital education design process, aiming to establish personalized and representative further education in dementia care applicable to various individual preferences and structural workplace challenges of healthcare professions.

Background: Despite recent advancements, medical technology has not yet reached its peak. Precision medicine is growing rapidly, thanks to machine learning breakthroughs powered by increased computational capabilities. This article explores a deep-learning application for computer-aided diagnosis in dermatology. Methods: Using a custom model based on EfficientNetB3 and deep learning, we propose an approach for skin lesion classification that offers superior results with smaller, cheaper, and faster inference times compared to other models. The skin images dataset used for this research includes 8,222 files selected from the authors' collection and the ISIC2019 archive, covering six dermatological conditions. Results: The model achieved 95.4% validation accuracy on four categories—melanoma, basal cell carcinoma, benign keratosis-like lesions, and melanocytic nevi—using an average of 1,600 images per category. Adding two categories with fewer images (about 700 each)—squamous cell carcinoma and actinic keratoses—reduced the validation accuracy to 88.8%. The model maintained accuracy on new clinical test images taken under the same conditions as the training dataset. Conclusions: The custom model demonstrated excellent performance on the diverse skin lesions dataset, with significant potential for further enhancements.

Colorectal cancer (CRC) is characterized by a pro-inflammatory microenvironment and features high energy supply molecules that assure tumor growth. A still underestimated macromolecule is the Inorganic Polyphosphate (iPolyP), a high-energy linear polymer that is ubiquitous in all forms of life. Made up of hundreds of repeated orthophosphate units, iPolyP is essential for a wide variety of functions in mammalian cells, including the regulation of proliferative signaling pathways. Some evidence has suggested its involvement in carcinogenesis, although more studies need to be pursued. In this study, we tested the role of iPolyP on CRC proliferation, using in-vitro and ex-vivo approaches, in order to evaluate its effect on tumor growth. We found that iPolyP is significantly increased in tumor tissues compared to the corresponding peritumoral counterparts. In addition, iPolyP signaling occurs through the TRPM8 receptor, promoting CRC cell proliferation. Pharmacological inhibition of TRPM8 or RNA interference experiments showed that the involvement of TRPM8 is essential, greater than that of the other two known iPolyP receptors, P2Y1 and RAGE. The presence of iPolyP drives cancer cells towards the mitotic phase of the cell cycle by enhancing the expression of ccnb1, which encodes for the Cyclin B protein. In-vitro 2D and 3D data reflected the ex-vivo results, obtained by the generation of CRC-derived organoids, which increased in size. These results indicate that iPolyP may be considered a novel and unexpected early biomarker supporting colorectal cancer cell proliferation.

Cancer cells are adept at reprogramming energy metabolism and the precise manifestation of this metabolic reprogramming exhibits heterogeneity across individuals (and from cell to cell). In this study, we analyzed the metabolic differences between interpersonal heterogeneous cancer phenotypes. We used divergence analysis on gene expression data of 1156 breast normal and tumor samples from The Cancer Genome Atlas (TCGA) and integrated this information with a genome-scale reconstruction of human metabolism to generate personalized, context-specific metabolic networks. Using this approach, we classified the samples into four distinct groups based on their metabolic profiles. Enrichment analysis of the subsystems indicated that amino acid metabolism, fatty acid oxidation, citric acid cycle, androgen and estrogen metabolism and ROS detoxification distinguished these four groups. Additionally, we developed a workflow to identify potential drugs that can selectively target genes associated with the reactions of interest. MG-132 (a proteasome inhibitor) and OSU-03012 (a celecoxib derivative) were the top-ranking drugs identified from our analysis and known to have anti-tumor activity. Our approach has the potential to provide mechanistic insights into cancer-specific metabolic dependencies, ultimately enabling the identification of potential drug targets for each patient independently, contributing to a rational personalized medicine approach.

Personalized medicine is an aspect of the P4 medicine (predictive, preventive, personalized and participatory) based precisely on the customization of all medical characters of each subject. In personalized medicine, the development of medical treatments and drugs is tailored to the individual characteristics and needs of each subject, according to the study of diseases at different scales from genotype to phenotype scale. To make concrete the goal of personalized medicine, it is necessary to employ high-throughput methodologies such as Next Generation Sequencing (NGS), Genome-Wide Association Studies (GWAS), Mass Spectrometry or Microarrays, that are able to investigate a single disease from a broader perspective. For example, by using genotyping microarrays (e.g. collections of Single Nucleotide Polymorphism - SNP) it is possible to uncover the reasons (i.e. mutation in genes) because a treatment works properly in some patients (for example absence of mutated genes), but it does not work (presence of mutated genes) in others. A side effect of high-throughput methodologies is the massive amount of data produced for each single experiment, that poses several challenges (e.g. high execution time and required memory) to bioinformatic software. Thus a main requirement of modern bioinformatic software is the use of good software engineering methods and efficient programming techniques, able to face those challenges, that include the use of parallel programming and efficient and compact data structures. Thus, to exploit all the potential of this massive amount of data in the short possible time (before that data becomes obsolete), the necessity to develop parallel software tools for efficient data collection and analysis arise. Moreover, due to the heterogeneity of the data produced by the different kinds of experimental platforms, it is necessary to automatize in a comprehensive software pipeline, the various steps that compose a bioinformatic analysis, such as: the preprocessing of raw data to remove noise or corrupted data; the annotation of data with external knowledge (e.g. Gene Ontology), and the integration of molecular data with clinical data. It should be noted that such steps are necessary to make statistical or data mining analysis more effective. This paper presents the design and the experimentation of a comprehensive software pipeline, named microPipe, for the preprocessing, annotation and analysis of microarray-based SNP genotyping data. A case study in pharmacogenomics is presented. The main advantages of using microPipe are: the reduction of errors that may happen when trying to make data compatible among different tools; the possibility to analyze in parallel huge datasets; the easy annotation and integration of data.

The Eurasian grapevine (Vitis vinifera L.) is the most widely grown horticultural crop in the world and is important for the economy of many countries. In the wine production chain, grape varieties play an important role as they directly influence the authenticity and classification of the product. Identifying the different grape varieties is therefore fundamental for quality control and inspection activities, as well as for regulating production. Currently, ampelography and molecular analysis are the main approaches to identifying grape varieties. However, both methods have limitations. Ampelography is subjective and prone to errors and is experiencing enormous difficulties as ampelographers are increasingly scarce. On the other hand, molecular analyses are very demanding in terms of cost and time. In this scenario, Deep Learning (DL) and Machine Learning (ML) methods have emerged as a classification alternative to deal with the scarcity of ampelographs and avoid molecular analyses. In this study, the most recent and current methods for identifying grapevine varieties using DL classification-based approaches are presented through a systematic literature review. The classification pipeline of the 31 studies found in the literature was described, highlighting its pros and cons. Most of the studies used DL-based models trained with leaf images acquired in a controlled environment at a maximum distance of 1.2 metres to classify grape varieties. In addition, there is a large gap between practical applications and the datasets used: a great lack of varieties, limited data acquired in the field and a lack of tests on plants under adverse conditions. Potential directions for improving this area of research were also presented.

Self-sustained learning (SSL) is considered a crucial factor in promoting students’ long-term learning and development in a sustainable society. To deepen the understanding of SSL, especially in the context of learning English as a second language, the current study examined the influence of intercultural communicative skills, language mindset, and positive L2 self on SSL in English learning among 1,238 college students in China. The results showed that students not only had a strong language mindset, positive L2 self, and intercultural communicative skills but also scored higher in long-term self-sustained English learning. Findings from the Structural Equation Modeling indicated that both language mindset and intercultural communicative skills positively predicted positive L2 self and self-sustained English learning. Furthermore, language mindset and intercultural communicative skills had an indirect impact on self-sustained English learning through positive L2 self. These findings suggest that to promote English learners’ SSL, instructors should help facilitate students’ intercultural communicative skills and foster a growth language mindset, as well as positive language learning beliefs.

Mass spectrometry (MS) has revolutionized clinical chemistry, offering unparalleled capabilities for biomolecule analysis. This review explores the growing significance of MS, particularly when coupled with liquid chromatography (LC-MS), in identifying disease biomarkers and quantifying biomolecules for diagnostic and prognostic purposes. The unique advantages of MS in accurately identifying and quantifying diverse molecules have positioned it as a cornerstone in personalized medicine advancement. MS-based technologies have transformed precision medicine, enabling comprehensive understanding of disease mechanisms and patient-specific treatment responses. LC-MS has shown exceptional utility in analyzing complex biological matrices, while high-resolution MS has expanded analytical capabilities, allowing detection of low-abundance molecules and elucidation of complex biological pathways. The integration of MS with other techniques, such as ion mobility spectrometry, has opened new avenues for biomarker discovery and validation. As we progress towards precision medicine, MS-based technologies will be crucial in addressing the challenges of individualized patient care, driving innovations in disease diagnosis, prognosis, and treatment strategies.

Tumor cells evolve in a complex and heterogeneous environment composed of different cell types and an extracellular matrix. Current 2D culture methods are very limited in their ability to mimic the cancer cell environment. In recent years, various 3D models of cancer cells have been developed, notably in the form of spheroids/organoids, using scaffold or cancer-on-chip devices. However, these models have the disadvantage of not being able to precisely control the organization of multiple cell types in complex architecture and are sometimes not very reproducible in their production, and this is especially true for spheroids. Three-dimensional bioprinting can produce complex, multi-cellular, and reproducible constructs in which the matrix composition and rigidity can be adapted locally or globally to the tumor model studied. For these reasons, 3D bioprinting seems to be the technique of choice to mimic the tumor microenvironment in vivo as closely as possible. In this review, we discuss different 3D-bioprinting technologies, including bioinks and crosslinkers that can be used for in vitro cancer models, and the techniques used to study cells grown in hydrogels; finally, we provide some applications of bioprinted cancer models.

The rapid expansion of Bangladesh's financial infrastructure, coupled with the increasing digitization of banking services, underscores the critical importance of cybersecurity. Machine learning and deep learning methodologies have emerged as potent tools in combating cyber threats such as intrusion, malware, and fraud. By harnessing artificial intelligence and data-driven analysis, these techniques enable real-time scrutiny of vast banking data, facilitating proactive defense against cyberattacks. Bangladesh's evolving financial landscape, characterized by governmental digitalization initiatives and a national financial inclusion strategy, heightens the imperative for robust cybersecurity measures. This paper explores the application of machine learning and deep learning techniques in bolstering banking cybersecurity within the context of Bangladesh's burgeoning digital economy.

The proliferation of fake news on social media platforms poses significant challenges to information authenticity and public trust. In this study, the impact of various word embedding techniques on the performance of both machine learning and deep learning models for fake news detection is systematically evaluated. The comprehensive TruthSeeker dataset, which spans over a decade of labeled news articles and social media posts, is utilized. Three popular word embedding techniques are analyzed: TF-IDF, Word2Vec, and FastText. These embeddings were tested with multiple machine learning classifiers, including logistic regression, naive Bayes, K-nearest neighbors, support vector machines, multilayer perceptrons, decision trees, and random forests, as well as deep learning models, specifically convolutional neural networks (CNNs). Among machine learning models, the Support Vector Machine (SVM) with TF-IDF achieved the highest accuracy of 99.03%, precision of 98.92%, recall of 99.15%, and F1 score of 99.03%. The Multilayer Perceptron (MLP) also demonstrated strong performance with Word2Vec, reaching an accuracy of 95.24%, precision of 94.95%, recall of 95.50%, and F1 score of 95.07%. For deep learning models, CNNs combined with TF-IDF embeddings showed outstanding results, achieving an accuracy of 98.99%, precision of 98.56%, recall of 99.48%, and F1 score of 99.02%. These results highlight the critical role of word embeddings in enhancing model accuracy and reliability in misinformation detection.

With advancements in digital technologies such as 5G communications, big data, and cloud computing, the components of network operation systems have become increasingly complex, significantly complicating system monitoring and maintenance. Correspondingly, automated log anomaly detection has become a crucial means to ensure stable network operation and protect them from malicious attacks or failures. Conventional machine learning and deep learning methods assume consistent distributions between the training and testing data, adhering to a closed-set recognition paradigm. Nevertheless, in realistic scenarios, systems may encounter new anomalies that were not present in the training data, especially in log anomaly detection. Inspired by evidential learning, we propose a novel anomaly detector called LogEDL, which supervises training of the model through an evidential loss function. Unlike traditional loss functions, the evidential loss function not only focuses on correct classification, but also quantifies the uncertainty of predictions. This enhances the robustness and accuracy of the model in handling anomaly detection tasks, while achieving functionality similar to open-set recognition. To evaluate the proposed LogEDL method, we conduct extensive experiments on three datasets, i.e., HDFS, BGL, and Thunderbird, to detect anomalous log sequences. Experimental results demonstrate that our proposed LogEDL achieves state-of-the-art performance in anomaly detection.

Regenerative endodontic procedures (REPs) aim at recreating dental pulp tissue using biomaterials such as hydrogels. Their bioactivity is mostly linked to the nature of biomolecules or chemical compounds that will compose the endodontic hydrogel. However, many other parameters, such as hydrogel concentration, type, bioactive molecules solubility, of apex size were reported to influence the hydrogel behavior, leading to the inability to predict the clinical outcome. This lack of knowledge and the various parameters that should be considered, suggest that biological activity of endodontic hydrogel is impossible to anticipate and could hinder the bench to bedside transition. We formulated in this article our opinion that most of these parameters could be identified, described, and studied. A second part of the document lists some challenges and perspectives, including development of future mathematical models able to explain and predict the bioactivity of these endodontic hydrogel in clinic.

Asthma is a prevalent disease that around 300 million worldwide, resulting in substantial morbidity, mortality, and economic burden on a global scale. New clinical and laboratory research has shed light on the immunology causing asthma. Asthma is now recognized as a heterogeneous disease. A personalized medicine is based on the classification of asthma by endotype by linking observable characteristics to immunological mechanisms. Identifying endotype mechanisms is essential for better characterizing patients and personalizing therapeutic approaches with novel biological agents that target specific immune pathways. This article provides a summary of the major immunological mechanisms involved in the pathogenesis and emergence of the disease's phenotypic features, as well as the individualized treatment for severe asthma subtypes

Discrete distribution estimation is a fundamental statistical tool, which is widely used to perform data analysis tasks in various applications involving sensitive personal information. Due to privacy concerns, individuals may not always provide their raw information, which leads to unpredictable biases in the final results of estimated distribution. Local Differential Privacy (LDP) is an advanced technique for privacy protection of discrete distribution estimation. Currently, typical LDP mechanisms provide same protection for all items in the domain, which imposes unnecessary perturbation on less sensitive items and thus degrades the utility of final results. Although, several recent works try to alleviate this problem, the utility can be further improved. In this paper, we propose a novel notion called Item-Oriented Personalized LDP (IPLDP), which independently perturbs different items with different privacy budgets to achieve personalized privacy protection. Furthermore, to satisfy IPLDP, we propose the Item-Oriented Personalized Randomized Response (IPRR) based on the observation that the sensitivity of data shows an inverse relationship with the population size of respective individuals. Theoretical analysis and experimental results demonstrate that our method can provide fine-grained privacy protection and improve data utility simultaneously.

The study was conducted on applying machine learning and data mining methods to personalizing the treatment. This allows investigating individual patient characteristics. Personalization is built on the clustering method and associative rules. It was suggested to determine the average distance between instances for optimal performance metrics finding. The formalization of the medical data pre-processing stage for finding personalized solutions based on current standards and pharmaceutical protocols is proposed. The model of patient data is built. The paper presents the novel approach to clustering built on ensemble of cluster algorithm with better than k-means algorithm Hopkins metrics. The personalized treatment usually is based on decision tree. Such approach requires a lot of computation time and cannot be paralyzed. Therefore, it is proposed to classify persons by conditions, to determine deviations of parameters from the normative parameters of the group, as well as the average parameters. This made it possible to create a personalized approach to treatment for each patient based on long-term monitoring. According to the results of the analysis, it becomes possible to predict the optimal conditions for a particular patient and to find the medicaments treatment according to personal characteristics.

Recent years have witnessed significant advances in reinforcement learning (RL), which has registered remarkable success in solving various sequential decision-making problems in machine learning. Most of the successful RL applications, e.g., the games of Go and Poker, robotics, and autonomous driving, involve the participation of more than one single agent, which naturally fall into the realm of multi-agent RL (MARL), a domain with a relatively long history, and has recently re-emerged due to advances in single-agent RL techniques. Although, typically, the communication protocol between agents is manually specified and not altered during training, recently, some papers have shown signs of trying to emerge a communication between agents on the one hand and, on the other hand, to understand what is exchanged between agents. So, there is a growing body of literature on this topic which includes qualitative and quantitative studies and the ones that apply mixed methods. This study presents the scoping review of the methodological strategies undertaken in a total of 16 research articles. The results present the critical appraisal of quantitative methods in terms of validity and reliability and for qualitative methods considering four trustworthiness factors. In the end, relevant insights are further explored with implications and reflections on how they can benefit one's research in the field.

The advancement of autonomous driving technology is becoming increasingly vital in the modern technological landscape, promising notable enhancements in safety, efficiency, traffic management, and energy use. Despite these benefits, conventional deep reinforcement learning algorithms often struggle to navigate complex driving environments effectively. To tackle this challenge, we propose a novel network called DynamicNoise, designed to significantly boost algorithmic performance by introducing noise into the Deep Q-Network (DQN) and Double Deep Q-Network (DDQN). Drawing inspiration from the NoiseNet architecture, DynamicNoise uses stochastic perturbations to improve the exploration capabilities of these models, leading to more robust learning outcomes. Our experiments demonstrate a 57.25% improvement in navigation effectiveness within a 2D experimental setting. Moreover, by integrating noise into the action selection and fully connected layers of the Soft Actor-Critic (SAC) model in the more complex 3D CARLA simulation environment, our approach achieved an 18.9% performance gain, substantially surpassing traditional methods. These results confirm that the DynamicNoise network significantly enhances the performance of autonomous driving systems across various simulated environments, regardless of their dimensionality and complexity, by improving their exploration capabilities rather than just their efficiency.

The success of near-infrared spectroscopy (NIRS) analysis hinges on the precision and robustness of the calibration model. Shallow learning (SL) algorithms like partial least squares discriminant analysis (PLS-DA) often fall short in capturing interrelationships between adjacent spectral variables, and the analysis results are easily affected by spectral noise, which dramatically limits the breadth and depth application of NIRS. Deep learning (DL) methods, with their capacity to discern intricate features from limited samples, are progressively integrated into NIRS. In this paper, two discriminant analysis problems including wheat kernels and Yali pears were used as examples, and several representative calibration models to research the robustness and effectiveness of the model. Additionally, this article proposed a near-infrared calibration model, which is based on gramian angular difference field and coordinate attention convolutional neural networks (G-CACNN). The research results show that, compared with SL, spectral preprocessing has a smaller impact on the analysis accuracy of consensus learning (CL) and DL, and the latter has the highest analysis accuracy in the modeling results of using the original spectrum. The accuracy of G-CACNN in two discrimination tasks is 98.48% and 99.39%, respectively. Finally, This research compared the performance of various models under noise to evaluate the robustness and noise resistance of the proposed method.

Multimodal remote sensing data classification can enhance the model’s ability to distinguish land features through multimodal data fusion. In this context, how to help models understand the relationship between multimodal data and target tasks is the focus of researchers. Inspired by human feedback learning mechanisms, causal reasoning mechanisms, and knowledge induction mechanisms, this paper integrates causal learning, reinforcement learning, and meta learning into a unified remote sensing data classification framework and proposed the causal meta-reinforcement learning (CMRL). First, based on feedback learning mechanisms, we have overcome the limitations of traditional implicit optimization of fusion features and customized a reinforcement learning environment for multimodal remote sensing data classification tasks. Through feedback interactive learning between agents and the environment, we help them understand the complex relationships between multimodal data and labels, thereby achieving full mining of multimodal complementary information. Second, based on the causal inference mechanism we designed causal distribution prediction action, classification reward, and causal intervention reward, capturing pure causal factors in multimodal data and cutting off false statistical associations between non-causal factors and class labels. Finally, based on the knowledge induction mechanism, we designed a bi-layer optimization mechanism based on meta-learning. By constructing a meta training task and meta validation task simulation model in the generalization scenario of unseen data, we helped the model induce cross-task shared knowledge, thereby improving its generalization ability for unseen multimodal data. The experimental results on multiple sets of multimodal datasets show that the proposed method achieved state-of-the-art performance in multimodal remote sensing data classification tasks.