13264 research outputs found
Sort by
Dual band polarization insensitive metamaterial absorber for EMI shielding from GSM and 5G communication systems
No full textIn this paper, a new metamaterial absorber (MMA) has been presented that shows dual-band absorption at 1.8 GHz and 3.5 GHz. The MMA cell has been designed on an FR4 substrate with the electrical dimension of 0.14λ0 × 0.14λ0 × 0.01λ0, calculating wavelength, λ0 at 1.8 GHz. Maximum absorption of 98.7% and 99.7% are attained by a unique design of a resonating patch consisting of two modified circular rings that are finalized through numerical Simulation in CST microwave studio. The MMA exhibits high angular stability up to 60° for incident angle as well as polarization angle variations. The analogous equivalent circuit is modeled in advanced design system (ADS) software, providing the same reflection, transmission, and absorption characteristics of 3D Simulation in CST. The absorption mechanism is investigated through current and electromagnetic field analysis. The MMA exhibits negative permittivity within 1 GHz—1.8 GHz and 2.08 GHz – 3.49 GHz and negative permeability with other frequency ranges. The prototype of a 3 × 6 array of the MMA cell is developed, and measurement is accomplished. The measured result exhibits well match with the simulated result. Moreover, The MMA displays good shielding effectiveness of 40.12 dB and 36.81 dB at 1.8 GHz and 3.5 GHz, respectively. The quality factors of the MMA are 30 and 22.3, with half power bandwidth of 60 MHz and 157 MHz. This new and unique MMA can be incorporated with various electronic devices for microwave shielding from GSM 1.8 GHz and sub-6, 5G 3.5 GHz signals
A wideband array antenna with a novel matching circuit and DGS structure for the sub 6 ghz applications
No full textThis paper proposes a novel matching network with a DGS structure to enhance the bandwidth of a
two-element array antenna. The array antenna only generates a single resonance with a conventional
partial ground plane with poor bandwidth and gain. Several dips are observed when the matching
network is loaded between the array elements, but their matching is not good enough. Then, the
partial ground is applied with a DGS structure, which improves the low-frequency band’s matching.
After that, three rectangular parasitic blocks are placed on the ground plane to improve the matching
of the high-frequency band. Parametric analysis is done on several parameters to position these two
bands close enough, considerably enhancing the bandwidth. The overall dimension of the proposed
antenna is 30 mm ×30 mm ×1.6 mm. A prototype of the proposed antenna is fabricated and measured.
Experimental results show that the antenna has an operating bandwidth of 70% for |S11| < −10 dB
ranging from 3.05GHz to 6.32 GHz, where a pick gain of 3.07 dBi is realized. The radiation patterns at
the two significant dips are stable. The cross-polarization level in the E plan is less than −16 dB at the
low resonant frequenc
An explainable AI-driven deep neural network for accurate breast cancer detection from histopathological and ultrasound images
No full textBreast cancer represents a significant global health challenge, which makes it essential to detect
breast cancer early and accurately to improve patient prognosis and reduce mortality rates. However,
traditional diagnostic processes relying on manual analysis of medical images are inherently complex
and subject to variability between observers, highlighting the urgent need for robust automated
breast cancer detection systems. While deep learning has demonstrated potential, many current
models struggle with limited accuracy and lack of interpretability. This research introduces the Deep
Neural Breast Cancer Detection (DNBCD) model, an explainable AI-based framework that utilizes deep
learning methods for classifying breast cancer using histopathological and ultrasound images. The
proposed model employs Densenet121 as a foundation, integrating customized Convolutional Neural
Network (CNN) layers including GlobalAveragePooling2D, Dense, and Dropout layers along with
transfer learning to achieve both high accuracy and interpretability for breast cancer diagnosis. The
proposed DNBCD model integrates several preprocessing techniques, including image normalization
and resizing, and augmentation techniques to enhance the model’s robustness and address class
imbalances using class weight. It employs Grad-CAM (Gradient-weighted Class Activation Mapping)
to offer visual justifications for its predictions, increasing trust and transparency among healthcare
providers. The model was assessed using two benchmark datasets: Breakhis-400x (B-400x) and Breast
Ultrasound Images Dataset (BUSI) containing 1820 and 1578 images, respectively. We systematically
divided the datasets into training (70%), testing (20%,) and validation (10%) sets, ensuring efficient
model training and evaluation obtaining accuracies of 93.97% for B-400x dataset having benign
and malignant classes and 89.87% for BUSI dataset having benign, malignant, and normal classes
for breast cancer detection. Experimental results demonstrate that the proposed DNBCD model
significantly outperforms existing state-of-the-art approaches with potential uses in clinical
environments. We also made all the materials publicly accessible for the research community at: https:/
/github.com/romzanalom/XAI-Based-Deep-Neural-Breast-Cancer-Detection
Consumer's Intention to Use Self-Service Technology to Do More with Less
No full textSelf-service technologies (SSTs) are electronic interfaces that let users generate services without the assistance of
a direct service representative. Many in-person service encounters are being replaced by self-service technology
in an effort to improve the accuracy, convenience, and speed of service transactions. SST is the most adopted
method and an alternative to traditional technology in this modern era. For example, a self-service laundry
shop that is open 24 hours a day, an ATM machine, and a robot that serves as a waiter are among the few SSTs
that are available to make our lives easier. This study aims to investigate consumers' intentions to use self-service
technology. A sample of 200 respondents is drawn using a simple random sampling technique. Four important
variables, namely perceived ease of use, perceived usefulness, brand image, and promotion stand as a basis for the
research framework and these variables are analysed using correlation and multiple regression analysis. The findings
from this research will help government and policymakers to assess the impact of consumer’s intention enabling
them to develop supportive policies and programs. These initiatives can enhance the marketing capabilities of
a company, to promote and contribute to the economic growth of the country. For instance, policymakers can
provide financial resources or training programs to small and medium-sized enterprises, empowering them to
invest in self-service technology to enhance their financial returns
Dataset on influence of individual entrepreneurial orientation, individual dynamic capabilities, and entrepreneurial self-efficacy on agropreneur performance
No full textAgricultural entrepreneurship, or better known as agropreneurship, is one of the key drivers of the agricultural sector’s growth in Malaysia. This dataset documents the traits that have an impact on the agropreneurs’ current financial performance. A total of 166 usable responses were collected, using an online questionnaire, from agropreneurs who attended the AGORA training organised by Malaysian Agricultural Research and Development Institute, started from May till November 2023. The online questionnaire includes demographic variables and traits that influence an agropreneur’s performance. The dataset is valuable to Malaysia government and agricultural agencies as well as future research seek to explore the personality and traits of successful youth agropreneurs
Design and Development of a Multifunctional Stepladder: Usability, Sustainability, and Cost-Effectiveness
No full textThis study presents the design, development, and evaluation of a multifunctional stepladder that integrates four functionalities: a stepladder, Pilates chair, wheelchair, and walking aid. Unlike existing research that focuses on single-function assistive devices, this study uniquely integrates a stepladder, wheelchair, walking aid, and Pilates chair into one multifunctional device, offering a compact, space-saving solution that addresses multiple daily needs in a single design. Building upon previous research, which conceptualized a multifunctional stepladder by synthesizing ideas, features, and functions from patent literature, existing products, and scientific articles, this study focuses on the design and testing phases to refine and validate the concept. Using sustainable materials like mild steel and aluminium, the design was optimized through structural simulations, ensuring durability under loads of up to 100 kg. Usability tests revealed that the invention significantly reduced task completion times, saved five times the space compared to single-function products, and provided enhanced versatility. Cost analysis highlighted its affordability, with a retail price of MYR 1392—approximately 35% lower than the combined cost of its single-function counterparts. Participant feedback noted strengths such as eco-friendliness, practicality, and ergonomic design, alongside areas for improvement, including portability, armrests, and storage. Future work includes enhanced portability for stair navigation, outdoor usability tests, and integration of smart technologies. This multifunctional stepladder significantly contributes to caregivers by reducing the physical burden of managing multiple assistive devices, enhancing efficiency in daily caregiving tasks, and providing a safer, more convenient tool that supports both mobility and exercise for elderly users. This multifunctional stepladder also offers a sustainable, cost-effective, and user-centric solution, addressing usability gaps while supporting global sustainability and accessibility initiatives
Decentralized EEG-based detection of major depressive disorder via transformer architectures and split learning
No full textIntroduction: Major Depressive Disorder (MDD) remains a critical mental health
concern, necessitating accurate detection. Traditional approaches to diagnosing
MDD often rely on manual Electroencephalography (EEG) analysis to identify
potential disorders. However, the inherent complexity of EEG signals along with
the human error in interpreting these readings requires the need for more
reliable, automated methods of detection.
Methods: This study utilizes EEG signals to classify MDD and healthy individuals
through a combination of machine learning, deep learning, and split learning
approaches. State of the art machine learning models i.e., Random Forest,
Support Vector Machine, and Gradient Boosting are utilized, while deep learning
models such as Transformers and Autoencoders are selected for their robust
feature-extraction capabilities. Traditional methods for training machine learning
and deep learning models raises data privacy concerns and require significant
computational resources. To address these issues, the study applies a split
learning framework. In this framework, an ensemble learning technique has been
utilized that combines the best performing machine and deep learning models.
Results: Results demonstrate a commendable classification performance with
certain ensemble methods, and a Transformer-Random Forest combination
achieved 99% accuracy. In addition, to address data-sharing constraints, a split
learning framework is implemented across three clients, yielding high accuracy
(over 95%) while preserving privacy. The best client recorded 96.23% accuracy,
underscoring the robustness of combining Transformers with Random Forest
under resource-constrained conditions.
Discussion: These findings demonstrate that distributed deep learning pipelines
can deliver precise MDD detection from EEG data without compromising
data security. Proposed framework keeps data on local nodes and only
exchanges intermediate representations. This approach meets institutional
privacy requirements while providing robust classification outcomes
Design of wearable social distancing card for enhancing public health safety compliance
No full textDuring the COVID-19 pandemic, physical distancing has been essential in preventing the virus's spread. The WHO has implemented physical distancing with at least one meter between individuals. Despite widespread awareness, people often forget to maintain a safe distance in public spaces. This paper addresses this issue by proposing an innovative solution: a wearable social distance card. The objective of this paper is to design a wearable device equipped with modern sensor technology to detect and promote safe distance practices. This device integrates ultrasonics and passive infrared sensors, an Arduino UNO, a buzzer, and an LCD display. The device uses high-frequency sound waves from ultrasonic sensors to measure the distance between individuals in real-time. When individuals reach the predetermined safe distance threshold, the buzzer emits a sound to notify both people to maintain a proper distance. A microcontroller-based wearable social distancing card helps the community practice social distancing. The prototype demonstrates accuracy in distance measurements, exhibiting an average deviation of ±4 cm. This wearable card aids in encouraging social distancing practices, supporting public health efforts to mitigate the spread of infectious diseases. The design represents a significant step forward in promoting social distancing compliance
Empowering minds: how self-efficacy, self-esteem, and social support drive digital mental health engagement
No full textIntroduction: Despite the rapid growth of research in digital health, there is a significant gap in understanding how psychological factors such as self-efficacy, self-esteem, and perceived social support collectively influence digital mental health engagement, particularly within the Malaysian context. While previous studies have explored these constructs individually, few have examined their integrated effects on user engagement with e-health platforms. The study aims to fill that gap by exploring the direct relationships between self-efficacy, self-esteem, and perceived social support with digital mental health engagement, while also analyzing the mediating role of perceived social support. The novelty of the research lies in integration of these psychological constructs into a unified conceptual framework to provide a more comprehensive understanding of digital mental health engagement. Methods: Using a quantitative, cross-sectional design, the study surveyed 400 active Malaysian e-health users through a self-administered questionnaire. The survey used validated scales to measure self-efficacy, self-esteem, perceived social support, and engagement with digital mental health platforms. Structural Equation Modeling (SEM) was employed to test five hypotheses regarding direct and mediated relationships. Results: Demographic analysis revealed that many participants were female (71.3%), aged between 25 and 45 years (76.6%), and from higher income brackets (RM5,001–RM20,000). WhatsApp (400 users) was found to be the most popular tool, followed by Facebook (387 users) and Instagram (371 users), highlighting the importance of these platforms in connectivity and information sharing. The study found that perceived social support had the strongest direct effect on digital mental health engagement (β = 0.523), followed by self-esteem (β = 0.384) and self-efficacy (β = 0.236). Additionally, perceived social support significantly mediated the impact of both self-efficacy and self-esteem on engagement. Discussion: These findings underscore the importance of fostering supportive digital environments to enhance users’ confidence and self-worth. However, the cross-sectional nature of the study limits causal interpretations, and the localized sample restricts generalizability. Future research should incorporate longitudinal methods and explore cultural differences. Overall, the study contributes to the development of effective digital mental health engagement strategies in Malaysia and beyond
Improving the Radiation Resistance of the Fiber via Bismuth Depositing
No full textErbium-doped fiber (EDF), bismuth-erbium co-doped fiber (BEDF) without bismuth deposited bismuth on the fiber surface (BEDF1), and BEDF with bismuth deposited on the fiber surface (BEDF2) were fabricated. Leveraging the Geant4 toolkit, the theoretical models of BEDFs are established. When the thickness of the bismuth deposited layer is 5~\mu m, the energy deposited in the core is reduced by 49.66% (5.91 MeV). The energy deposited in the core is reduced by 19.7% (1.06 MeV) when Bi3+ ions’ concentration is 10 wt%. EDF, BEDF1, and BEDF2 were irradiated with 0.3-, 0.5-, 0.8-, and 1.5-kGy doses of Co60 source. After radiation with 1.5 kGy, radiation-induced absorption (RIA) of BEDF2 at 1300 nm is 20.31% (1.01 dB/m) lower than that of BEDF1. The radiation induced gain variation (RIGV) and fluorescence lifetime of EDF are decreased with the increase in irradiation dose, while the RIGV and lifetime of BEDF1 and BEDF2 both increased initially and then decreased linearly. The normalized gain of BEDF2 is increased by up to 1.08 dB/m after 0.5-kGy radiation. Bismuth ions in the fiber core improve the radiation resistance of the fiber, and those deposited on the fiber surface further improve the radiation resistance. The research results have reference value and application potential in radiation environment detection and space satellite positioning