Resource-constrained Environments Research Articles

Optimizing network lifetime: ERBS-REE for resilient object detection and tracking in resource-constrained WSN environments

Optimizing network lifetime: ERBS-REE for resilient object detection and tracking in resource-constrained WSN environments

The community of inquiry as a tool for measuring student engagement in blended massive open online courses (MOOCs): a case study of university students in a developing country

While massive open online courses (MOOCs) promise to democratise access to education, the literature reveals a nuanced understanding of engagement in these settings, especially in resource-constrained environments. Blended MOOCs combine MOOCs and physical classroom settings of contents and instructions. This study extends this discourse by focusing on blended MOOCs, which remain under-explored in the context of developing countries. The blended MOOC at the University of Cape Coast (UCC), Ghana, deals with third-party MOOCs as open educational resources (OERs) integrated with campus-based courses. UCC students have been using such blended MOOCs since 2016, when all level 100 students were mandated to enrol in a course entitled Information Technology Skills (ITS101). ITS101 is aligned to courses in a MOOC platform called Alison as an OER. Students' engagement is key to their continued use and satisfaction with online learning, such as MOOCs. However, among all the e-learning modes, students' engagement is the lowest in MOOCs, leading to high dropout rates. Blended MOOCs are one of the techniques recommended to reverse the undesirability of MOOCs, including engagement. However, few studies have been conducted on students' engagement in blended MOOCs, especially among university students in sub-Saharan Africa using MOOCs as OERs. Thus, this paper aims to measure student engagement in blended MOOCs using the revised Community of Inquiry for university students in a developing country. The rationale is to determine whether factors affect engagement positively or negatively. A two-stage cluster sampling technique was used to determine the participants for this study. A list of blended MOOC classes offered at UCC was obtained from the staff's mailing list. In the first stage, academic levels (100, 200, 300 and 800) were randomly selected from the strata using a lottery sampling technique. In the second stage, another simple random selection of blended MOOC courses or classes was used within each selected academic level. All students in the selected classes were then included in the study. Partial Least Squares Structural Equation Modelling was used to validate the model on the predictive relationships existing among the four presences (cognitive, learning, social and teaching) and engagement. Results from the structural model analysis proved a statistically significant predictive relationship among the constructs within the model. Learning presence had the most significant effect on student engagement. Thus, it should be included as one of the presences in the community of inquiry.

Collaborative Localization Strategy Based on Node Selection and Power Allocation in Resource-Constrained Environments

Collaborative Localization Strategy Based on Node Selection and Power Allocation in Resource-Constrained Environments

Stoma-related complications and quality-of-life assessment: A cross-sectional study with patients from Ethiopia and Sweden.

The situation for patients with ostomy can be challenging, probably more in a resource-constrained environment. Our objective was to evaluate quality of life (QoL) (using EQ5D-5L) and stoma-specific QoL (using Stoma QoL) in a high- and low-income setting. In this cross-sectional study from the Tikur Anbessa Specialized Hospital (TASH), Addis Ababa, and South General Hospital (SGH), Stockholm, patients with a permanent or temporary ostomy at TASH (EthioPerm), (EthioTemp), and patients with ostomy at SGH (SweSto) were included in October 2022-January 2023. Patients N=66 were included in groups: EthioPerm N=28, EthioTemp N=17, and SweSto N=21. In EthioTemp, 88% used homemade stoma bags. Although morbidity related to the nipple itself was similar in the groups, the overall score from Stoma QoL was significantly lower in EthioPerm, 48/100 than in SweSto, 74/100. Scores were significantly lower for pouch-related problems and social interactions in Ethiopian patients. In EthioPerm, 71% of the patients worried that they were a burden to the people close to them compared to 14% in SweSto (p<0.001). Leakage was over four times more common in EthioPerm than in SweSto. Mean overall EQ5D-5L score was 18 percentage points lower than the national mean score in EthioPerm and 2 percentage points lower in SweSto. QoL was more affected in the Ethiopian study participants than in the Swedish, even when commercial stoma bags were available. The largest problems were leakage, embarrassment with social interactions, and pouch-related problems. NCT05970458 Clinicaltrials.gov, https://clinicaltrials.gov/study/NCT05970458?locStr=Ethiopia&country=Ethiopia&distance=50&cond=Stoma%20Ileostomy&rank=1.

STAM-LSGRU: a spatiotemporal radar echo extrapolation algorithm with edge computing for short-term forecasting

With the advent of Mobile Edge Computing (MEC), shifting data processing from cloud centers to the network edge presents an advanced computational paradigm for addressing latency-sensitive applications. Specifically, in radar systems, the real-time processing and prediction of radar echo data pose significant challenges in dynamic and resource-constrained environments. MEC, by processing data near its source, not only significantly reduces communication latency and enhances bandwidth utilization but also diminishes the necessity of transmitting large volumes of data to the cloud, which is crucial for improving the timeliness and efficiency of radar data processing. To meet this demand, this paper proposes a model that integrates a spatiotemporal Attention Module (STAM) with a Long Short-Term Memory Gated Recurrent Unit (ST-ConvLSGRU) to enhance the accuracy of radar echo prediction while leveraging the advantages of MEC. STAM, by extending the spatiotemporal receptive field of the prediction units, effectively captures key inter-frame motion information, while optimizations to the convolutional structure and loss function further boost the model’s predictive performance. Experimental results demonstrate that our approach significantly improves the accuracy of short-term weather forecasting in a mobile edge computing environment, showcasing an efficient and practical solution for processing radar echo data under dynamic, resource-limited conditions.

Three-way interaction effect of entrepreneurial orientation, CEO power, and organizational slack on Indonesian firms’ performance

This study aims to explore the complex interplay between entrepreneurial orientation, CEO power, and organizational slack, and their collective impact on firm performance within Indonesia’s manufacturing sector, providing actionable insights to optimize operations in a dynamic, resource-constrained environment. The paper employs a longitudinal approach, utilizing dynamic panel data from 127 publicly-traded manufacturing companies listed on the Indonesia Stock Exchange from 2014 to 2022, focusing on this sector due to its significant role in Indonesia’s economy and the unique challenges these firms face in adapting to market changes and competitive pressures. The study initially found that entrepreneurial orientation did not significantly influence firm performance. However, when CEO power was introduced into the analysis, it significantly, albeit negatively, moderated the effect of entrepreneurial orientation, suggesting that higher CEO power may actually diminish beneficial impacts of entrepreneurial orientation on performance (β = –48.041, p &amp;amp;lt; 0.05). Importantly, the analysis revealed that organizational slack can positively interact with both entrepreneurial orientation and CEO power, mitigating this negative influence and enhancing firm performance (β = 15.261, p &amp;amp;lt; 0.05). These findings illuminate the complex interdependencies within strategic management, underscoring the necessity of aligning upper echelon power dynamics with organizational resources. This alignment is crucial for leveraging the full potential of entrepreneurial orientation to enhance firm performance. AcknowledgmentsThis study received sponsorship from Beasiswa Pendidikan Indonesia and Lembaga Pengelola Dana Pendidikan (Indonesia Endowment Fund for Education). The support from these organizations has been invaluable in the completion of this work, enabling comprehensive data collection and analysis crucial to complete the study.

A rehearsal framework for computational efficiency in online continual learning

AbstractIn the realm of online continual learning, models are expected to adapt to an ever-changing environment. One of the most persistent hurdles in this adaptation is the mitigation of a phenomenon called "Catastrophic Forgetting" (CF). This critical condition occurs when models trained on non-identically distributed data lose performance in previously learned tasks. Rehearsal methods, leveraging the ability to replay older samples, aim to address this challenge by incorporating a buffer of past training samples. However, the absence of known task boundaries complicates the adaptation of current CF mitigation methods. This paper proposes a method attuned to data stream characteristics and online model performance in a resource-constrained environment. The number of training iterations and learning rate emerges as crucial hyperparameters, impacting the efficacy and efficiency of online continual learning. Up to this point, we propose a combination of Experience Replay methodologies, a Drift Detector, and various training convergence policies, specially tailored for scenarios with unknown task boundaries. Experimental results demonstrate the effectiveness of our approach, maintaining or enhancing performance compared to baseline methods, while significantly improving computational efficiency.

Lightweight Deep Learning for Resource-Constrained Environments: A Survey

Over the past decade, the dominance of deep learning has prevailed across various domains of artificial intelligence, including natural language processing, computer vision, and biomedical signal processing. While there have been remarkable improvements in model accuracy, deploying these models on lightweight devices, such as mobile phones and microcontrollers, is constrained by limited resources. In this survey, we provide comprehensive design guidance tailored for these devices, detailing the meticulous design of lightweight models, compression methods, and hardware acceleration strategies. The principal goal of this work is to explore methods and concepts for getting around hardware constraints without compromising the model’s accuracy. Additionally, we explore two notable paths for lightweight deep learning in the future: deployment techniques for TinyML and Large Language Models. Although these paths undoubtedly have potential, they also present significant challenges, encouraging research into unexplored areas.

Lightweight Convolutional Network with Integrated Attention Mechanism for Missing Bolt Detection in Railways

Railway infrastructure safety is a paramount concern, with bolt integrity being a critical component. In the realm of railway maintenance, the detection of missing bolts is a vital task that ensures the stability and safety of tracks. Traditionally, this task has been approached through manual inspections or conventional automated methods, which are often time-consuming, costly, and prone to human error. Addressing these challenges, this paper presents a state-of-the-art solution with the development of a lightweight convolutional neural network (CNN) featuring an integrated attention mechanism. This novel model is engineered to be computationally efficient while maintaining high accuracy, making it particularly suitable for real-time analysis in resource-constrained environments commonly found in railway inspections. The proposed CNN utilises a distinctive architecture that synergises the speed of lightweight networks with the precision of attention-based mechanisms. By integrating an attention mechanism, the network selectively concentrates on regions of interest within the image, effectively enhancing the model’s capability to identify missing bolts with remarkable accuracy. Comprehensive testing showcases a remarkable 96.43% accuracy and an impressive 96 F1-score, substantially outperforming existing deep learning frameworks in the context of missing bolt detection. Key contributions of this research include the model’s innovative attention-integrated approach, which significantly reduces the model complexity without compromising detection performance. Additionally, the model offers scalability and adaptability to various railway settings, proving its efficacy not just in controlled environments but also in diverse real-world scenarios. Extensive experiments, rigorous evaluations, and real-time deployment results collectively underscore the transformative potential of the presented CNN model in advancing the domain of railway safety maintenance.

Age-of-Information based Multi-UAV Trajectories Using Deep Reinforcement Learning

The use of unmanned aerial vehicles (UAVs) as a unified platform for sensing and communication is especially relevant in environments with inadequate infrastructure. In this paper, a multi-UAV system is constructed for dynamic data collection in a resource-constrained environment. The devised approach involves the implementation of an access platform referred to as the Access UAV (A_UAV). This A_UAV orchestrates the data collection process from Inspection-UAVs (I_UAVs), each equipped with a visual sensor, facilitating the relay of collected data to the cloud. Our approach jointly considers the trajectory scheduling of A_UAV and I_UAV to collect data from specific points in a decentralized manner. Specifically, a Deep Reinforcement Learning framework utilizing an actor-critic network is formulated for A_UAV, aiming to generate an equitable access schedule for I_UAVs. Moreover, trajectory scheduling of A_UAV ensures dynamic data collection while minimizing total system energy and the Age of Information (AoI) of data arriving from I_UAVs. The simulation results validate the performance of our proposed approach against several baselines under different parameter settings.

A mobile edge computing-focused transferable sensitive data identification method based on product quantization

Sensitive data identification represents the initial and crucial step in safeguarding sensitive information. With the ongoing evolution of the industrial internet, including its interconnectivity across various sectors like the electric power industry, the potential for sensitive data to traverse different domains increases, thereby altering the composition of sensitive data. Consequently, traditional approaches reliant on sensitive vocabularies struggle to adequately address the challenges posed by identifying sensitive data in the era of information abundance. Drawing inspiration from advancements in natural language processing within the realm of deep learning, we propose a transferable Sensitive Data Identification method based on Product Quantization, named PQ-SDI. This innovative approach harnesses both the composition and contextual cues within textual data to accurately pinpoint sensitive information within the context of Mobile Edge Computing (MEC). Notably, PQ-SDI exhibits proficiency not only within a singular domain but also demonstrates adaptability to new domains following training on heterogeneous datasets. Moreover, the method autonomously identifies sensitive data throughout the entire process, eliminating the necessity for human upkeep of sensitive vocabularies. Extensive experimentation with the PQ-SDI model across four real-world datasets, resulting in performance improvements ranging from 2% to 5% over the baseline model and achieves an accuracy of up to 94.41%. In cross-domain trials, PQ-SDI achieved comparable accuracy to training and identification within the same domain. Furthermore, our experiments showcased the product quantization technique significantly reduces the parameter size by tens of times for the subsequent sensitive data identification phase, particularly beneficial for resource-constrained environments characteristic of MEC scenarios. This inherent advantage not only bolsters sensitive data protection but also mitigates the risk of data leakage during transmission, thus enhancing overall security measures in MEC environments.

Task Scheduling in a Cloud Environment: A comparative Study

In this study, we explore the application of Monarch Butterfly Optimization (MBO) algorithms for task scheduling in cloud computing, comparing its performance against widely used optimization techniques, namely Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO).Task scheduling in the cloud is a critical aspect influencing resource utilization, turnaround time, and overall system efficiency. MBO, known for its effective exploration- exploitation balance, is examined for its suitability in addressing the complexities of cloud computing environments. The study investigates MBO's advantages, such as enhanced adaptability to dynamic conditions, effective handling of multi-objective optimization, and its consideration of bandwidth as a critical resource. Comparative analyses with ACO and PSO highlight MBO's superior performance in achieving near-optimal task schedules, emphasizing its potential to offer innovative solutions to the challenges posed by task scheduling in dynamic and resource-constrained cloud environments. This research contributes valuable insights into the strengths of MBO, paving the way for advancements in optimization methodologies tailored for cloud computing systems.

Developing classroom ICT teaching techniques, principles and practice for teachers in rural Ghana without access to computers or internet: a framework based on literature review

PurposeThis paper aims to propose a conceptual framework for integrating information and communication technologies (ICT) into teaching practices for educators in rural Ghana lacking computer and internet access. Bridging the rural–urban digital divide is critical for equitable quality education. This research identifies distinct infrastructure, financial, human, and socio-cultural challenges of ICT adoption in underserved Ghanaian schools through an exhaustive literature review. The framework provides tailored strategies to empower teachers to leverage ICT for innovative pedagogy despite limitations. Ultimately, the goal is to offer pragmatic, evidence-based solutions to advance ICT-enabled instruction and digital equity in marginalized settings.Design/methodology/approachThis study utilizes a qualitative literature review methodology to develop a conceptual framework for ICT teaching techniques tailored to the rural Ghanaian context. Published empirical studies, theoretical papers, and reports focused on ICT education in developing countries, particularly sub-Saharan Africa, were systematically reviewed across multidisciplinary databases. Relevant theories and models, like TPACK and SAMR, were analyzed to determine appropriate modifications for the framework design. The proposed strategies emphasize affordable technologies, teacher training, localized content, and culturally relevant pedagogies selected based on evidence from literature to address the unique barriers and needs of the rural setting.FindingsThe literature review revealed profound infrastructure, financial, human, and socio-cultural barriers to ICT adoption in rural Ghanaian schools. These include lack of electricity, internet, hardware, and software alongside inadequate teacher training and support. The proposed conceptual framework offers solutions including leveraging low-cost technologies like mobile devices, prioritizing teacher professional development, developing localized digitally-enabled resources, and fostering community participation. Culturally relevant pedagogies are also emphasized. This tailored, evidence-based approach holds promise for enhancing ICT-enabled instruction and bridging digital divides to promote equitable quality education. Further findings will emerge through framework implementation and evaluation.Research limitations/implicationsThis conceptual framework requires implementation and evaluation to truly determine its impact and effectiveness in the intended rural Ghanaian context. As a literature review, primary data collection was not undertaken. Practical application of the framework in schools can yield direct insights through surveys, interviews, observations, and analysis of key educational outcomes pre and post-implementation. Comparisons with control groups may further illuminate its efficacy. Monitoring and adaptation over time is also essential. Ultimately, this research sets the foundation for praxis-oriented work transforming ICT integration in marginalized communities to advance digital equity and inclusion.Practical implicationsThis research provides an evidence-based, context-aware framework with pragmatic solutions to enhance ICT-enabled teaching and learning in underserved rural Ghanaian schools. It offers guidance to teachers, administrators, and policymakers on integrating technology despite infrastructure and resource constraints. The emphasis on cost-effective innovations, educator training, localized content, and community participation can inform planning and investments in ICT education. Broader applications include adapting the framework for other developing country contexts facing comparable challenges. Overall, this work promotes digital equity and quality education access, catalyzing social development and aligning with the UN Sustainable Development Goals.Social implicationsThis research highlights the urgent need to address digital divides exacerbating educational inequalities between rural and urban communities in Ghana. Lack of technology access and utilization in rural schools threatens to widen gaps in skills, capabilities, and opportunities for students. Strategic integration of ICT can promote social inclusion, cultivate digital literacy, and equip youth to participate in the global digital economy, thereby enhancing prospects. More broadly, advancing digital equity and ICT-enabled education empowers marginalized groups, fosters community empowerment, and drives progressive social change. This work underscores that localized, context-appropriate solutions are pivotal for promoting development and social justice.Originality/valueThis research fills a crucial gap by proposing a conceptual framework tailored specifically for integrating ICT in teaching practices in rural Ghanaian schools lacking digital access, responding to context-specific barriers and needs. While extensive literature examines ICT education models globally, few studies offer localized strategies for resource-constrained environments. This framework adapts established models based on a systematic literature review, emphasizing affordability, capacity building, and cultural relevance. By synthesizing evidence-based, pragmatic solutions to advance ICT-enabled instruction in marginalized settings, this work provides a valuable foundation for praxis transforming rural education. The localization approach could inform frameworks for other developing regions.

Leveraging machine learning for vaccine distribution in resource-limited settings: A synthesis of approaches

Vaccine distribution in resource-limited settings remains a crucial global health challenge, exacerbated by factors such as inadequate infrastructure, limited resources, and complex supply chains. Leveraging machine learning (ML) holds promise for optimizing distribution efficiency and ensuring equitable access to life-saving vaccines. This paper synthesizes various ML approaches aimed at addressing vaccine distribution challenges in resource-constrained environments. The literature review examines existing research on ML applications in healthcare and vaccine distribution, highlighting key findings and methodologies. Methodologically, criteria were established for selecting relevant studies, with a focus on ML techniques and their effectiveness in resource-limited contexts. Key ML approaches identified include predictive analytics for demand forecasting, route optimization algorithms for efficient vaccine delivery, and decision support systems for prioritizing distribution efforts. Case studies illustrate successful ML implementations in real-world settings, showcasing improved vaccine coverage and reduced wastage. Despite promising results, challenges persist, including data scarcity, model generalization, and ethical considerations. Future research directions include enhancing data collection methods, refining ML algorithms for specific contexts, and integrating ML solutions into existing healthcare systems. In conclusion, this synthesis underscores the transformative potential of ML in revolutionizing vaccine distribution in resource-limited settings. By addressing logistical barriers and optimizing resource allocation, ML-driven approaches offer a pathway towards achieving universal immunization coverage and mitigating the impact of infectious diseases on vulnerable populations. Keywords: Machine Learning, Vaccine Distribution, Resource-Limited Settings, Synthesis of Approaches.

Calibration of simulation model to analyze hospital building energy performance

This research paper addresses the calibration of an energy simulation model for a naturally ventilated hospital building located in Mangalore, Karnataka, India, in a warm, humid climatic zone. The study aims to enhance the accuracy of energy consumption calculations through a calibration process by developing the energy model of a reference hospital building using EnergyPlus software. The required architectural design data, construction details, occupancy and schedule data and services are collected through documents, energy audits and documentation, measured drawings, and semi-structured interviews. Calibration parameters are systematically identified and adjusted through an iterative process. The monthly electricity bill is used to validate the simulation model. The simulation model reached during the validation process has an excellent Coefficient of Variation-Root Mean Squared Error of 4.24% and a Normalized Mean Bias Error of −4.42%, both of which meet the ASHRAE-approved accuracy standards. The paper also discusses the challenges and emphasizes the need for a nuanced approach to modeling healthcare facilities. Notwithstanding the challenges faced, the study offers insightful information about how to calibrate simulation models for hospital energy usage with the significant influence of scheduling for artificial lighting and equipment usage. By pioneering a manual calibration approach tailored to hospital simulation models in the warm humid climate of India, this paper offers a novel and practical solution to address the challenges of energy performance analysis in resource-constrained environments. The calibrated simulation model presented in this study is a valuable tool for assessing and improving the energy performance of naturally ventilated healthcare facilities. By creating alternative space layouts, the research findings aim to foster the development of hospital infrastructure that is both ecologically friendly and energy-efficient in a specific climatic context. This research is assured to make a substantial contribution to the advancement of sustainable hospital design in warm-humid climates, with implications for both academia and industry.

OpenCell: A low-cost, open-source, 3-in-1 device for DNA extraction.

High-cost DNA extraction procedures pose significant challenges for budget-constrained laboratories. To address this, we introduce OpenCell, an economical, open-source, 3-in-1 laboratory device that combines the functionalities of a bead homogenizer, a microcentrifuge, and a vortex mixer. OpenCell utilizes modular attachments that magnetically connect to a central rotating brushless motor. This motor couples to an epicyclic gearing mechanism, enabling efficient bead homogenization, vortex mixing, and centrifugation within one compact unit. OpenCell's design incorporates multiple redundant safety features, ensuring both the device's and operator's safety. Additional features such as RPM measurement, programmable timers, battery operation, and optional speed control make OpenCell a reliable and reproducible laboratory instrument. In our study, OpenCell successfully isolated DNA from Spinacia oleracea (spinach), with an average yield of 2.3 μg and an A260/A280 ratio of 1.77, demonstrating its effectiveness for downstream applications such as Polymerase Chain Reaction (PCR) amplification. With its compact size (20 cm x 28 cm x 6.7 cm) and lightweight design (0.8 kg), comparable to the size and weight of a laptop, OpenCell is portable, making it an attractive component of a 'lab-in-a-backpack' for resource-constrained environments in low-and-middle-income countries and synthetic biology in remote field stations. Leveraging the accessibility of 3D printing and off-the-shelf components, OpenCell can be manufactured and assembled at a low unit cost of less than $50, providing an affordable alternative to expensive laboratory equipment costing over $4000. OpenCell aims to overcome the barriers to entry in synthetic biology research and contribute to the growing collection of frugal and open hardware.

COMPUTATIONAL INTELLIGENCE BASED ROUTING SOLUTIONS FOR MOBILE AD HOC NETWORKS

Mobile Ad Hoc Networks (MANETs) refers to a wireless network where nodes dynamically establish and maintain connections with each other, directly without the need for a centralized infrastructure. MANETs are characterized by their decentralized nature, presenting challenges for routing protocols concerning rapid data transmission, minimized delay and Quality of Service (QoS) provisioning. Traditional routing protocols may struggle to perform optimally in MANETs due to frequent changes in network topology and resource constraints. The development of intelligent routing protocols tailored to the specific characteristics of MANETs is vital for dynamic and resource-constrained environments. Computational intelligence (CI) techniques have gained attention for designing intelligent routing protocols for MANETs. This paper provides a comprehensive overview of majorly used CI methodologies in MANETs. The paper reviews existing research on intelligent routing protocols based on CI techniques in MANETs and identifies future research directions in this field to address the unique challenges of MANETs and improve network performance and reliability.

ELIDS: Ensemble Feature Selection for Lightweight IDS against DDoS attacks in resource-constrained IoT environment

ELIDS: Ensemble Feature Selection for Lightweight IDS against DDoS attacks in resource-constrained IoT environment

Longitudinal Parallel Compression Suture To Control PPH Due to Placenta Previa &amp; Accreta.

Objective: To determine the efficacy and safety of longitudinal parallel compression sutures to control postpartum haemorrhage (PPH) in placenta previa/accreta patients. Methodology: This cross-sectional study was conducted at Department of obstetrics and gynaecology, Peoples University of Medical &amp; Health Sciences for Women; Nawab shah for a period of six months from 7th June 2023 to 6th Dec 2023, after getting approval from ERB of People University of Medical &amp; Health Science. Sample size achieved through non-probability consecutive sampling. Women fulfilling inclusion criteria were selected from the operation theatre of Obstetrics &amp; Gynaecology, PUMHS Nawabshah. To control life threatening PPH owing to placenta previa (with or without accreta) during caesarean section, longitudinal parallel compression sutures in operating room. Outcome variables observed includes amount of blood loss, number of days stayed in the hospital, need for concurrent procedure, need of hysterectomy, and maternal death. Data was entered and analysed on SPSS 25. Results: The stitch was applied in 31 patients. The technique was effective, and successful as a single procedure in 11 patients (35.43%). 6 patients (19%) needed additional tamponade; 12 patients (38%) also received concurrent uterine artery ligation while 2 needed ovarian artery ligation. Despite concurrent procedure 3 patients ended up for hysterectomy (9.1%). The most common indication of caesarean section was placenta previa (51%), mostly done at a gestational age of 37 weeks ± 1.84 standard deviation. Mean age of the patient was 29 years. Estimated blood loss was more than 1.5 litre as seen in 22.6% of the patients and 71.6% patients need blood transfusion. Uterine atony was a common cause of PPH (51%). 51.6% of patients discharged within 3 days of surgery. Two patients expired. Conclusion: The innovative suture techniques have shown promise in improving surgical outcomes in resource-constrained environments, especially in developing countries. This effectiveness can help bridge the gap between limited resources and the need for quality surgical care. Key words: PPH, Uterine Compression Suture, Placenta Previa, Placenta Accreta.

WaveSegNet: An Efficient Method for Scrap Steel Segmentation Utilizing Wavelet Transform and Multiscale Focusing

Scrap steel represents a sustainable and recyclable resource, instrumental in diminishing carbon footprints and facilitating the eco-friendly evolution of the steel sector. However, current scrap steel recycling faces a series of challenges, such as high labor intensity and occupational risks for inspectors, complex and diverse sources of scrap steel, varying types of materials, and difficulties in quantifying and standardizing manual visual inspection and rating. Specifically, we propose WaveSegNet, which is based on wavelet transform and a multiscale focusing structure for scrap steel segmentation. Firstly, we utilize wavelet transform to process images and extract features at different frequencies to capture details and structural information in the images. Secondly, we introduce a mechanism of multiscale focusing to further enhance the accuracy of segmentation by extracting and perceiving features at different scales. Through experiments conducted on the public Cityscapes dataset and scrap steel datasets, we have found that WaveSegNet consistently demonstrates superior performance, achieving the highest scores on the mIoU metric. Particularly notable is its performance on the real-world scrap steel dataset, where it outperforms other segmentation algorithms with an average increase of 3.98% in mIoU(SS), reaching 69.8%, and a significant boost of nearly 5.98% in mIoU(MS), achieving 74.8%. These results underscore WaveSegNet’s exceptional capabilities in processing scrap steel images. Additionally, on the publicly available Cityscapes dataset, WaveSegNet shows notable performance enhancements compared with the next best model, Segformer. Moreover, with its modest parameters and computational demands (34.1 M and 322 GFLOPs), WaveSegNet proves to be an ideal choice for resource-constrained environments, demonstrating high computational efficiency and broad applicability. These experimental results attest to the immense potential of WaveSegNet in intelligent scrap steel rating and provide a new solution for the scrap steel recycling industry. These experimental results attest to the immense potential of WaveSegNet in intelligent scrap steel rating and provide a new solution for the scrap steel recycling industry.