Collaborative Training Research Articles

FL-W3S: Cross-domain federated learning for weakly supervised semantic segmentation of white blood cells.

Segmentation models for clinical data experience severe performance degradation when trained on a single client from one domain and distributed to other clients from different domain. Federated Learning (FL) provides a solution by enabling multi-party collaborative learning without compromising the confidentiality of clients' private data. In this paper, we propose a cross-domain FL method for Weakly Supervised Semantic Segmentation (FL-W3S) of white blood cells in microscopic images. We perform model training on multiple clients with different data distributions to obtain a global aggregated model using only image-level class labels for semantic segmentation of white blood cells. A multi-class token transformer model learns the relationship between patch tokens and class tokens during collaborative learning and generates class-specific localization maps for mask predictions. To rectify the localization maps, we use patch-level pairwise affinity obtained from patch-to-patch transformer attention. We evaluate performance of the proposed semantic segmentation method on two different datasets of white blood cells from different domains. Our experimental results show that for two datasets, there is 2.56% and 1.39% increase in performance of the proposed method over existing state-of-the-art methods. The combination of federated learning for collaborative model training while preserving data privacy, alongside white blood cell segmentation techniques for precise cell identification, enhances diagnostic accuracy and personalized treatment strategies in clinical applications, particularly in hematology and pathology. More specifically, it involves isolating white blood cell from blood smear for further analysis such as automated blood cell counting, morphological analysis, cell classification, disease diagnosis and monitoring.

Genomic sequencing: the case for equity of care in the era of personalized medicine.

Over the past two decades, genomic sequencing (exome and genome) has proven to be critical in providing a faster and more accurate diagnosis as well as tailored treatment plans for a variety of populations. Despite its potential, disparities in access to genomic sequencing persist, predominantly among underrepresented and socioeconomically disadvantaged groups and populations. This inequity stems from factors such as: 1) high costs of sequencing, 2) significant gaps in insurance coverage, 3) limited availability of genetic services in many healthcare institutions and geographic areas, and 4) lack of diversity in genetic research and databases. Addressing these barriers is essential to realizing the full benefits of personalized treatment approaches for all individuals. By doing so, healthcare systems can move towards a more inclusive model that delivers optimal care for everyone. This manuscript emphasizes these issues by considering diverse perspectives from various ethnic groups, summarizing findings across different patient populations (adults, pediatrics, critical/non-critical care), and highlighting the importance of collaboration and workforce training in genomic sequencing and interpretation. IMPACT: Presentation of exemplary studies demonstrating the advantages of genomic sequencing in various clinical settings and a variety of high-risk populations. Review of obstacles in providing equitable genomic medical care and the importance of national and international collaborations An overview of the ethical aspects of genomic sequencing is provided.

Survey of pharmacists' perspectives on preparing discharge medicine lists: a cross-sectional single-centre study.

This study explored South-East Queensland Australian pharmacists' perspectives on preparing discharge medicine lists, specifically involvement of pharmacy assistants, use of electronic medication management software, and expanding pharmacists' scope during discharge. Electronic survey distributed to pharmacists during December 2021 and data collected over 3 weeks. Pharmacists supported increased involvement of pharmacy assistants (with structured collaborative training), pharmacist-led medication reconciliation, and producing the discharge medicine list directly from the electronic record. Increased scope of pharmacy assistants to complete technical tasks will improve pharmacists' capacity to integrate with the medical team and perform discharge medication reconciliation.

Multi-proposal collaboration and multi-task training for weakly-supervised video moment retrieval

Multi-proposal collaboration and multi-task training for weakly-supervised video moment retrieval

Utilizing ATLS® Trauma Management Protocols: The Gold Standard for Safe and Effective Care.

Introduction: Approximately 5.8 million people die each year as a result of injuries. Traumatic injury is the leading cause of death under the age of 45 in the U.S. and worldwide. Traumatic injuries represent a significant global health burden, demanding swift and decisive intervention to mitigate morbidity and mortality. The Advanced Trauma Life Support (ATLS) course has emerged as a cornerstone in trauma management, offering healthcare providers a systematic approach to assessment, resuscitation, and definitive care. This article explores the pivotal role of ATLS protocols as the gold standard for safe and effective trauma management. ATLS protocols prioritize a structured assessment methodology, employing the ABCDE approach to rapidly identify and address life-threatening injuries. By adhering to this systematic framework, healthcare providers can promptly initiate time-critical interventions, minimizing the risk of adverse outcomes. Moreover, ATLS protocols promote standardization of care, ensuring consistent management practices across diverse clinical settings. Through multidisciplinary collaboration and continuous training, healthcare teams are equipped to deliver coordinated and efficient care, optimizing patient outcomes. In conclusion, using ATLS trauma management protocols represents more than a guideline; it embodies a commitment to excellence in patient care. By embracing the principles of ATLS, healthcare providers uphold the gold standard for safe, effective, and compassionate trauma management, ultimately saving lives and restoring hope in times of crisis.

Out-of-Distribution Detection via outlier exposure in federated learning.

Among various out-of-distribution (OOD) detection methods in neural networks, outlier exposure (OE) using auxiliary data has shown to achieve practical performance. However, existing OE methods are typically assumed to run in a centralized manner, and thus are not feasible for a standard federated learning (FL) setting where each client has low computing power and cannot collect a variety of auxiliary samples. To address this issue, we propose a practical yet realistic OE scenario in FL where only the central server has a large amount of outlier data and a relatively small amount of in-distribution (ID) data is given to each client. For this scenario, we introduce an effective OE-based OOD detection method, called internal separation & backstage collaboration, which makes the best use of many auxiliary outlier samples without sacrificing the ultimate goal of FL, that is, privacy preservation as well as collaborative training performance. The most challenging part is how to make the same effect in our scenario as in joint centralized training with outliers and ID samples. Our main strategy (internal separation) is to jointly train the feature vectors of an internal layer with outliers in the back layers of the global model, while ensuring privacy preservation. We also suggest an collaborative approach (backstage collaboration) where multiple back layers are trained together to detect OOD samples. Our extensive experiments demonstrate that our method shows remarkable detection performance, compared to baseline approaches in the proposed OE scenario.

Combining Immersive Simulation with a Collaborative Procedural Training on Local Anesthetic Systemic Toxicity and Fascia Iliaca Compartment Block: A Pilot Study

Introduction: Readiness to perform a wide variety of procedures or manage nearly any patient presentation remains an essential aspect of emergency medicine training and practice. Often, simulation is needed to supplement real-life exposure to provide comfort and knowledge, particularly with rarer pathology and procedures. As the scope of practice continues to grow, newer procedures, such as ultrasound (US)-guided nerve blocks (UGNB), are becoming integrated into resident training, building on previously established skills. The fascia iliaca compartment block (FICB) is performed on patients with specific femoral fractures and is a now a component of standard multimodal pain regimens, with US-guidance limiting adverse events. Given the need for high volumes of local anesthetic to perform the block it is imperative for clinicians to understand dosing as well as recognize and treat local anesthetic systemic toxicity (LAST). With sparse literature on sequential immersive and procedural simulation involving intertwined topics, this presents a unique opportunity for learners. Methods: To study the perceived knowledge and comfort with FICB and LAST, a pilot study was developed with two separate but concurrent one-hour simulations completed encompassing one of each topic over one day. We surveyed 19 learners, consisting of residents ranging from postgraduate years 1–3, prior to and immediately following completion, regarding their perceptions. We used the Stuart-Maxwell test to compare survey data. Results: More than half of participants (56%) had not received prior formal training on FICB. There was a positive trend in perceived confidence and knowledge with visualizing relevant anatomy (4.0 [2.0–6.0] vs 9.0 [7.5–10.0], P = 0.10), performing FICB (4.0 [1.0–5.0] vs 9.0 [7.0–10.0, P = 0.08]), and perceived ability to teach their peers (3.0 [1.0–5.0] vs 8.5 [7.0–10.0], P = 0.20). Perceived ability in diagnosing and managing LAST also increased following the simulation (5.0 [3.0–6.0] vs 6.0 [6.0–7.0], P = 0.12 and 3.0 [2.0–6.0] vs 6.0 [6.0–7.0], P = 0.08, respectively). Conclusion: Learners’ perceptions of this simulation experience echo the findings of previous studies in which simulation can be used to teach procedures and pathology; of note, however, we presented a novel experience with a combination of immersive and procedural simulation.

VFL-Cafe: Communication-Efficient Vertical Federated Learning via Dynamic Caching and Feature Selection.

Vertical Federated Learning (VFL) is a promising category of Federated Learning that enables collaborative model training among distributed parties with data privacy protection. Due to its unique training architecture, a key challenge of VFL is high communication cost due to transmitting intermediate results between the Active Party and Passive Parties. Current communication-efficient VFL methods rely on using stale results without meticulous selection, which can impair model accuracy, particularly in noisy data environments. To address these limitations, this work proposes VFL-Cafe, a new VFL training method that leverages dynamic caching and feature selection to boost communication efficiency and model accuracy. In each communication round, the employed caching scheme allows multiple batches of intermediate results to be cached and strategically reused by different parties, reducing the communication overhead while maintaining model accuracy. Additionally, to eliminate the negative impact of noisy features that may undermine the effectiveness of using stale results to reduce communication rounds and incur significant model degradation, a feature selection strategy is integrated into each round of local updates. Theoretical analysis is then conducted to provide guidance on cache configuration, optimizing performance. Finally, extensive experimental results validate VFL-Cafe's efficacy, demonstrating remarkable improvements in communication efficiency and model accuracy.

Decentralized Federated Learning with Prototype Exchange

As AI applications become increasingly integrated into daily life, protecting user privacy while enabling collaborative model training has become a crucial challenge, especially in decentralized edge computing environments. Traditional federated learning (FL) approaches, which rely on centralized model aggregation, struggle in such settings due to bandwidth limitations, data heterogeneity, and varying device capabilities among edge nodes. To address these issues, we propose PearFL, a decentralized FL framework that enhances collaboration and model generalization by introducing prototype exchange mechanisms. PearFL allows each client to share lightweight prototype information with its neighbors, minimizing communication overhead and improving model consistency across distributed devices. Experimental evaluations on benchmark datasets, including MNIST, CIFAR-10, and CIFAR-100, demonstrate that PearFL achieves superior communication efficiency, convergence speed, and accuracy compared to conventional FL methods. These results confirm PearFL’s efficacy as a scalable solution for decentralized learning in heterogeneous and resource-constrained environments.

Evaluating the implementation of online research training and mentorship among early-career family physicians in sub-Saharan Africa.

Research is needed to improve the performance of primary health care. In Africa, few family physicians conduct research, and therefore an online research training and mentorship programme was developed to build research capacity amongst novice and early career researchers. To evaluate the implementation of the AfriWon Research Collaborative (ARC) training and e-mentorship programme in sub-Saharan Africa. A 10-module online curriculum was supported by peer and faculty e-mentorship, to mentor participants in writing a research protocol. A convergent mixed methods study combined quantitative and qualitative data to evaluate nine implementation outcomes. Fifty-three participants (20 mentees, 19 peer mentors, and 14 faculty mentors), mostly male (70%), participated in the ARC online programme. The programme was seen as an acceptable and appropriate initiative. Mentees were mostly postgraduate students from African countries. Faculty mentors were mostly experienced researchers from outside of Africa. There were issues with team selection, orientation, communication, and role clarification. Only 35% of the mentees completed the programme. Alignment of mentoring in teams and engagement with the online learning materials was an issue. Costs were relatively modest and dependent on donor funds. Despite many challenges, the majority of participants supported the sustainability of the programme. The evaluation highlights the strengths and weaknesses of the ARC programme and e-mentoring. The ARC working group needed to ensure better organization and leadership of the teams. Going forward the programme should focus more on developing peer mentors and local supervisory capacity as well as the mentees.

Toward secure industrial internet of behaviours: a federated learning-based lightweight human behaviour recognition method with selective state space models

Human behaviour recognition is one of the most fundamental tasks in Industrial Internet of Behaviour (IIoB) and is crucial for the safe and reliable IIoB. Existing methods lacks adaptability and transferability. In addition, there is a data isolation problem among different users. Therefore, there is an urgent requirement to construct a secure and adaptive human behaviour recognition model in IIoB without violating the privacy of users. Mamba, a structured state space model that integrates a selection mechanism and a scan module, is used for time series modelling tasks. To tackle the aforementioned problems, an Federated Learning-based lightweight human behaviour recognition model with selective state space models is proposed. First, we design a human behaviour recognition model integrating Mamba and residual structure to achieve lightweight and secure human behaviour modelling. In addition, considering data privacy and training efficiency, a decentralised dynamic FL framework is designed to achieve lightweight and secure model collaborative training, including: initial selection of source users, model aggregation strategy based on dynamic weighting, and fine-tuning module based on small-sample data, to improve the training efficiency of the model and the accuracy of human behaviour recognition. Extensive experiments are conducted to prove the superior performance of the proposed method.

Patient and Public Involvement in Pharmacoepidemiological Research: An Environmental Scan of an Emerging Area.

Patient and public involvement (PPI) in research is required to improve the relevance, feasibility, and interpretability. However, research on PPIs in pharmacoepidemiology (PE) is still limited. This study aimed to provide an overview of PPI implementation in pharmacoepidemiology through an environmental scan. The environmental scan combined systematic reviews and expert interviews. A systematic review was conducted in MEDLINE, EMBASE, and the Cochrane Library from January 1, 2010, to June 30, 2023, to identify PE studies in which PPIs were mentioned. An additional review covered British Medical Journal's (BMJ) original articles from January 1, 2019, to June 30, 2023, via a similar method. In parallel, a cross-sectional study was conducted with a standardized questionnaire for French PE research teams, involving interviews via videoconference. We identified 3615 references for screening, among which 232 were selected for full-text screening. However, no studies have reported the use of PPIs in PE studies. The additional BMJ review identified 1058 references, 74 of which met the full-text selection criteria, and eight were included. Of 13 French PE research teams surveyed, three had prior PPI experience, and 12 affirmed the relevance of PPI in PE. The respondents identified barriers such as PE's complexity (n = 9). They suggested training for patients (n = 9) and collaboration with specialist teams (n = 5) to facilitate PPI. Our environmental scan highlighted the emergence and relevance of PPIs in PE studies, even though they are still uncommon. Tools are needed to acculture and assist PE researchers in engaging in PPIs.

DualRec: A Collaborative Training Framework for Device and Cloud Recommendation Models

DualRec: A Collaborative Training Framework for Device and Cloud Recommendation Models

Collaborative 360° virtual reality training of medical students in clinical examinations

ABSTRACT Simulation-based training in computer-generated environments has always played an important role in clinical medical education. Recently, there has been a growing interest in using 360° videos of real-life situations for training in health professions. Several studies report positive results from using 360° Virtual Reality for individuals, yet there are currently no studies on collaborative 360° Virtual Reality training. In this paper, we evaluate how 360° Virtual Reality can support collaborative training in clinical medical education. The study population consisted of 14 medical students in semester 5 of their Bachelor’s programme. The students were divided into three groups before watching and annotating a 360° video of an authentic learning situation inside a collaborative immersive virtual reality space. The original video shows a problem-based examination of the collateral and cruciate ligaments of the knee performed by students under the supervision of a professor. After training in collaborative 360° Virtual Reality, students then had to perform the same tests in a physical examination. The students’ performance was subsequently evaluated by a professor with expertise in knee examinations. The results show that 12 out of 14 students received a score of 2 for one or more tests, thereby meeting the required learning objective. One student received a score of 1 and one student did not perform any of the tests. The students actively use the tools provided by the software and different communicative strategies when working collaboratively in 360° Virtual Reality, which enables them to perform the tests in the physical examination by transferring their constructed knowledge. The results indicate that our pedagogical design in collaborative immersive 360° Virtual Reality can become a relevant addition to face-to-face clinical medical training.

Distributed Database Architectures for Federated Medical Training

Federated Learning (FL) has emerged as a promising approach for collaborative medical model training while preserving patient privacy. This research proposes the integration of FL with distributed database architectures to enable secure and efficient medical model training across diverse healthcare institutions. The approach addresses challenges such as real-time data synchronization, data heterogeneity, and low-latency model updates. Key innovations include hybrid SQL/NoSQL databases for structured and unstructured data, dynamic partitioning for improved data locality, and adaptive indexing for optimized query performance. The system incorporates secure data handling mechanisms like encryption and differential privacy, ensuring compliance with healthcare regulations. Scalability is achieved through decentralized database management, enabling broad healthcare node participation. The framework’s effectiveness is evaluated in real-world smart healthcare networks, focusing on model accuracy, query latency, scalability, and energy efficiency, with potential impacts on personalized medicine and collaborative healthcare analytics.

Empowering small businesses through innovative marketing technologies: A blueprint for growth and competitiveness

Small businesses are the backbone of the global economy but often face significant challenges in competing with larger enterprises, particularly in the digital age. This publication explores how innovative marketing technologies can empower small businesses to overcome these challenges, drive growth, and enhance customer engagement. By leveraging tools such as artificial intelligence (AI), customer relationship management (CRM) systems, social media analytics, and personalized marketing platforms, small businesses can transform their marketing strategies and achieve competitive advantage. The discussion begins with a broad overview of the evolving marketing landscape, highlighting the increasing importance of data-driven decision-making and digital-first approaches. It then narrows its focus to the specific needs of small businesses, emphasizing the cost-effectiveness and scalability of modern marketing tools. Case studies of successful implementations illustrate how technologies like AI-powered chatbots, programmatic advertising, and predictive analytics can streamline operations, improve customer experiences, and maximize ROI. Moreover, the publication examines the role of marketing professionals as strategic partners in guiding small businesses through the adoption of these technologies. It addresses potential barriers, such as limited budgets and technical expertise, and offers practical solutions, including collaborative training programs and access to open-source tools. Ultimately, this blueprint positions marketing innovation as a catalyst for small business success, advocating for a balanced approach that aligns technological adoption with core business values. The publication concludes by outlining actionable strategies for marketing professionals to empower small businesses, ensuring their sustained growth and competitiveness in an ever-evolving marketplace.

Privacy-Preserving Federated Learning-Based Intrusion Detection System for IoHT Devices

In recent years, Internet of Healthcare Things (IoHT) devices have attracted significant attention from computer scientists, healthcare professionals, and patients. These devices enable patients, especially in areas without access to hospitals, to easily record and transmit their health data to medical staff via the Internet. However, the analysis of sensitive health information necessitates a secure environment to safeguard patient privacy. Given the sensitivity of healthcare data, ensuring security and privacy is crucial in this sector. Federated learning (FL) provides a solution by enabling collaborative model training without sharing sensitive health data with third parties. Despite FL addressing some privacy concerns, the privacy of IoHT data remains an area needing further development. In this paper, we propose a privacy-preserving federated learning framework to enhance the privacy of IoHT data. Our approach integrates federated learning with ϵ-differential privacy to design an effective and secure intrusion detection system (IDS) for identifying cyberattacks on the network traffic of IoHT devices. In our FL-based framework, SECIoHT-FL, we employ deep neural network (DNN) including convolutional neural network (CNN) models. We assess the performance of the SECIoHT-FL framework using metrics such as accuracy, precision, recall, F1-score, and privacy budget (ϵ). The results confirm the efficacy and efficiency of the framework. For instance, the proposed CNN model within SECIoHT-FL achieved an accuracy of 95.48% and a privacy budget (ϵ) of 0.34 when detecting attacks on one of the datasets used in the experiments. To facilitate the understanding of the models and the reproduction of the experiments, we provide the explainability of the results by using SHAP and share the source code of the framework publicly as free and open-source software.

Prototype‐Based Collaborative Learning in UAV‐Assisted Edge Computing Networks

ABSTRACTContextThe rise of artificial intelligence of things (AloT) has enabled smart cities and industries, and UAV‐assisted edge computing networks are an important technology to support the above scenarios. UAV‐assisted refers to leveraging UAVs as a dynamic, flexible infrastructure to assist edge network data processing and communication tasks. Multiple UAVs can use their own resources, and collaborate edge servers to train artificial intelligence (Al) models.ObjectiveCompared with cloud‐based collaborative computing scenarios, UAV‐assisted edge collaborative learning can reduce training and inference delays and improve user satisfaction. However, UAV‐assisted edge networks scenario brings new challenges in terms of transmission burden and energy consumption.MethodThis paper proposes a prototype‐based joint optimization and training software system. The system consists of an optimization module and a training module. The optimization module first models an optimization problem including energy consumption and prototype error. Then it solves the optimization problem by problem transformation and plans the location of each UAV given the objects' position. After UAVs fly to the designated area and complete data collection, UAVs and the edge server train a model according to the proposed prototype‐based collaborative training module. Our training module enables multiple UAVs and an edge server to collaboratively train a model by lightweight prototype transmission and prototype aggregation. We also prove the convergence of the proposed collaborative training method.ResultsResults show our method reduces prototype error and energy consumption by at least 12.31% and improves model accuracy by 3.62% with a little communication burden.ConclusionFinally, we verify system performance through experiments.

EdgeGuard: Decentralized Medical Resource Orchestration via Blockchain-Secured Federated Learning in IoMT Networks

The development of medical data and resources has become essential for enhancing patient outcomes and operational efficiency in an age when digital innovation in healthcare is becoming more important. The rapid growth of the Internet of Medical Things (IoMT) is changing healthcare data management, but it also brings serious issues like data privacy, malicious attacks, and service quality. In this study, we present EdgeGuard, a novel decentralized architecture that combines blockchain technology, federated learning, and edge computing to address those challenges and coordinate medical resources across IoMT networks. EdgeGuard uses a privacy-preserving federated learning approach to keep sensitive medical data local and to promote collaborative model training, solving essential issues. To prevent data modification and unauthorized access, it uses a blockchain-based access control and integrity verification system. EdgeGuard uses edge computing to improve system scalability and efficiency by offloading computational tasks from IoMT devices with limited resources. We have made several technological advances, including a lightweight blockchain consensus mechanism designed for IoMT networks, an adaptive edge resource allocation method based on reinforcement learning, and a federated learning algorithm optimized for medical data with differential privacy. We also create an access control system based on smart contracts and a secure multi-party computing protocol for model updates. EdgeGuard outperforms existing solutions in terms of computational performance, data value, and privacy protection across a wide range of real-world medical datasets. This work enhances safe, effective, and privacy-preserving medical data management in IoMT ecosystems while maintaining outstanding standards for data security and resource efficiency, enabling large-scale collaborative learning in healthcare.

An End-Cloud Collaborative Federated Learning Debugging Framework for Data Heterogeneity

End-cloud collaborative computing framework ensures the security and privacy of edge device data, enabling collaborative training of global models without direct data exchange. However, in practical scenarios, anomalies in training or edge device data may severely degrade or disable the global model’s performance. Existing frameworks lack effective debugging and anomaly localization, hindering real-time monitoring and precise identification of abnormal edge devices in data heterogeneity scenarios. In this paper, we propose a new method named FedCheck, a debugging framework for end-cloud collaborative federated learning that enables real-time alerts and detects abnormal devices for nonindependent and identically distributed (nonIID) data without disrupting the regular training process. Specifically, we employ a model similarity-based method to quantitatively assess the degree of device anomaly in data heterogeneity scenarios, supporting real-time alerts during the end-cloud collaboration process. Furthermore, a simulation program replays the training process based on recorded telemetry data, facilitating backtracking debugging of any training round and the status of edge devices. Finally, the framework removes abnormal devices and repairs the global model. Experiments on MNIST and Fashion-MNIST datasets demonstrate that FedCheck can effectively detect and locate abnormal devices in data heterogeneity scenarios. Even in large-scale federated learning, it maintains high detection performance and exhibits good scalability.