A pilot study of measuring emotional response and perception of LLM-generated questionnaire and human-generated questionnaires.

The advent of ChatGPT has sparked a heated debate surrounding natural language processing technology and AI-powered chatbots, leading to extensive research and applications across various disciplines. This pilot study aims to investigate the impact of ChatGPT on users' experiences by administering two distinct questionnaires, one generated by humans and the other by ChatGPT, along with an Emotion Detecting Model. A total of 14 participants (7 female and 7 male) aged between 18 and 35 years were recruited, resulting in the collection of 8672 ChatGPT-associated data points and 8797 human-associated data points. Data analysis was conducted using Analysis of Variance (ANOVA). The results indicate that the utilization of ChatGPT enhances participants' happiness levels and reduces their sadness levels. While no significant gender influences were observed, variations were found about specific emotions. It is important to note that the limited sample size, narrow age range, and potential cultural impacts restrict the generalizability of the findings to a broader population. Future research directions should explore the impact of incorporating additional language models or chatbots on user emotions, particularly among specific age groups such as older individuals and teenagers. As one of the pioneering works evaluating the human perception of ChatGPT text and communication, it is noteworthy that ChatGPT received positive evaluations and demonstrated effectiveness in generating extensive questionnaires.

An Observational Pilot Study of a Tailored Environmental Monitoring and Alert System for Improved Management of Chronic Respiratory Diseases.

Objective: Chronic lung-related diseases, with asthma being the most prominent example, characterized by diverse symptoms and triggers, present significant challenges in disease management and prediction of exacerbations across patients. This research aimed to devise a practical solution by introducing a personalized alert system tailored to individual lung function and environmental conditions, offering a holistic approach for the management of a range of chronic respiratory conditions. Methods: In response to these challenges, we developed a personalized alert system based on individual lung function tests conducted in diverse environmental conditions, as determined by air-quality sensors. Our research was substantiated through an observational pilot study involving twelve healthy participants. These participants were exposed to varying air quality, temperature, and humidity conditions, and their lung function, as indicated by peak expiratory flow (PEF) values, was monitored. Results: The study revealed pronounced variability in pulmonary responses across different environments. Leveraging these findings, we proposed a design of a personalized alarm system that monitors air quality in real-time and issues alerts under potentially unfavorable environmental conditions. Additionally, we investigated the use of basic machine learning techniques to predict PEF values in these varied environmental settings. Discussion: The proposed system offers a proactive approach for individuals, particularly those with asthma, to actively manage their respiratory health. By providing real-time monitoring and personalized alerts, it aims to minimize exposure to potential asthma triggers. Ultimately, our system seeks to empower individuals with the tools for timely intervention, potentially reducing discomfort and enhancing management of asthma symptoms.

Node embedding-based graph autoencoder outlier detection for adverse pregnancy outcomes.

Adverse pregnancy outcomes, such as low birth weight (LBW) and preterm birth (PTB), can have serious consequences for both the mother and infant. Early prediction of such outcomes is important for their prevention. Previous studies using traditional machine learning (ML) models for predicting PTB and LBW have encountered two important limitations: extreme class imbalance in medical datasets and the inability to account for complex relational structures between entities. To address these limitations, we propose a node embedding-based graph outlier detection algorithm to predict adverse pregnancy outcomes. We developed a knowledge graph using a well-curated representative dataset of the Emirati population and two node embedding algorithms. The graph autoencoder (GAE) was trained by applying a combination of original risk factors and node embedding features. Samples that were difficult to reconstruct at the output of GAE were identified as outliers considered representing PTB and LBW samples. Our experiments using LBW, PTB, and very PTB datasets demonstrated that incorporating node embedding considerably improved performance, achieving a 12% higher AUC-ROC compared to traditional GAE. Our study demonstrates the effectiveness of node embedding and graph outlier detection in improving the prediction performance of adverse pregnancy outcomes in well-curated population datasets.

Infant Low Birth Weight Prediction Using Graph Embedding Features.

Low Birth weight (LBW) infants pose a serious public health concern worldwide in both the short and long term for infants and their mothers. Infant weight prediction prior to birth can help to identify risk factors and reduce the risk of infant morbidity and mortality. Although many Machine Learning (ML) algorithms have been proposed for LBW prediction using maternal features and produced considerable model performance, their performance needs to be improved so that they can be adapted in real-world clinical settings. Existing algorithms used for LBW classification often fail to capture structural information from the tabular dataset of patients with different complications. Therefore, to improve the LBW classification performance, we propose a solution by transforming the tabular data into a knowledge graph with the aim that patients from the same class (normal or LBW) exhibit similar patterns in the graphs. To achieve this, several features related to each node are extracted such as node embedding using node2vec algorithm, node degree, node similarity, nearest neighbors, etc. Our method is evaluated on a real-life dataset obtained from a large cohort study in the United Arab Emirates which contains data from 3453 patients. Multiple experiments were performed using the seven most commonly used ML models on the original dataset, graph features, and a combination of features, respectively. Experimental results show that our proposed method achieved the best performance with an area under the curve of 0.834 which is over 6% improvement compared to using the original risk factors without transforming them into knowledge graphs. Furthermore, we provide the clinical relevance of the proposed model that are important for the model to be adapted in clinical settings.

Development of YOLOv5-Based Real-Time Smart Monitoring System for Increasing Lab Safety Awareness in Educational Institutions.

The term "smart lab" refers to a system that provides a novel and flexible approach to automating and connecting current laboratory processes. In education, laboratory safety is an essential component of undergraduate laboratory classes. The institution provides formal training for the students working in the labs that involve potential exposure to a wide range of hazards, including chemical, biological, and physical agents. During the laboratory safety lessons, the instructor explains the lab safety protocols and the use of personal protective equipment (PPE) to prevent unwanted accidents. However, it is not always guaranteed that students follow safety procedures throughout all lab sessions. Currently, the lab supervisors monitor the use of PPE, which is time consuming, laborious, and impossible to see each student. Consequently, students may unintentionally commit unrecognizable unsafe acts, which can lead to unwanted situations. Therefore, the aim of the research article was to propose a real-time smart vision-based lab-safety monitoring system to verify the PPE compliance of students, i.e., whether the student is wearing a mask, gloves, lab coat, and goggles, from image/video in real time. The YOLOv5 (YOLOv5l, YOLOv5m, YOLOv5n, YOLOv5s, and YOLOv5x) and YOLOv7 models were trained using a self-created novel dataset named SLS (Students Lab Safety). The dataset comprises four classes, namely, gloves, helmets, masks, and goggles, and 481 images, having a resolution of 835 × 1000, acquired from various research laboratories of the United Arab Emirates University. The performance of the different YOLOv5 and YOLOv7 versions is compared based on instances' size using evaluation metrics such as precision, F1 score, recall, and mAP (mean average precision). The experimental results demonstrated that all the models showed promising performance in detecting PPE in educational labs. The YOLOv5n approach achieved the highest mAP of 77.40% for small and large instances, followed by the YOLOv5m model having a mAP of 75.30%. A report detailing each student's PPE compliance in the lab can be prepared based on data collected in real time and stored in the proposed system. Overall, the proposed approach can be utilized to make laboratories smarter by enhancing the efficacy of safety in research settings; this, in turn, will aid the students in establishing a health and safety culture among students.

Real-Time Non-Contact Millimeter Wave Radar-Based Vital Sign Detection.

In this paper, the extraction of the life activity spectrum based on the millimeter (mm) wave radar is designed to realize the detection of target objects and the threshold trigger module. The maximum likelihood estimation method is selected to complete the design of the average early warning probability trigger function. The threshold trigger module is designed for the echo signal of static objects in the echo signal. It will interfere with the extraction of Doppler frequency shift results. The moving target detection method is selected, and the filter is designed. The static clutter interference is filtered without affecting the phase difference between the detection sequences, and the highlight target signal is improved. The frequency and displacement of thoracic movement are used as the detection data. Through the Fourier transform calculation of the sequence, the spectrum value is extracted within the estimated range of the heartbeat and respiration spectrum, and the heartbeat and respiration signals are picked up. The proposed design uses Modelsim and Quartus for CO-simulation to complete the simulation verification of the function, extract the number of logical units occupied by computing resources, and verify the algorithm with the vital signs experiment. The heartbeat and respiration were detected using the sports bracelet; the relative errors of heartbeat detection were 0-6.3%, the respiration detection was 0-9.5%, and the relative errors of heartbeat detection were overwhelmingly less than 5%.

Exfoliation of MoS2 Quantum Dots: Recent Progress and Challenges.

Although, quantum dots (QDs) of two-dimensional (2D) molybdenum disulfide (MoS2) have shown great potential for various applications, such as sensing, catalysis, energy storage, and electronics. However, the lack of a simple, scalable, and inexpensive fabrication method for QDs is still a challenge. To overcome this challenge, a lot of attention has been given to the fabrication of QDs, and several fabrication strategies have been established. These exfoliation processes are mainly divided into two categories, the 'top-down' and 'bottom-up' methods. In this review, we have discussed different top-down exfoliation methods used for the fabrication of MoS2 QDs and the advantages and limitations of these methods. A detailed description of the various properties of QDs is also presented.

Infant birth weight estimation and low birth weight classification in United Arab Emirates using machine learning algorithms.

Accurate prediction of a newborn's birth weight (BW) is a crucial determinant to evaluate the newborn's health and safety. Infants with low BW (LBW) are at a higher risk of serious short- and long-term health outcomes. Over the past decade, machine learning (ML) techniques have shown a successful breakthrough in the field of medical diagnostics. Various automated systems have been proposed that use maternal features for LBW prediction. However, each proposed system uses different maternal features for LBW classification and estimation. Therefore, this paper provides a detailed setup for BW estimation and LBW classification. Multiple subsets of features were combined to perform predictions with and without feature selection techniques. Furthermore, the synthetic minority oversampling technique was employed to oversample the minority class. The performance of 30 ML algorithms was evaluated for both infant BW estimation and LBW classification. Experiments were performed on a self-created dataset with 88 features. The dataset was obtained from 821 women from three hospitals in the United Arab Emirates. Different performance metrics, such as mean absolute error and mean absolute percent error, were used for BW estimation. Accuracy, precision, recall, F-scores, and confusion matrices were used for LBW classification. Extensive experiments performed using five-folds cross validation show that the best weight estimation was obtained using Random Forest algorithm with mean absolute error of 294.53 g while the best classification performance was obtained using Logistic Regression with SMOTE oversampling techniques that achieved accuracy, precision, recall and F1 score of 90.24%, 87.6%, 90.2% and 0.89, respectively. The results also suggest that features such as diabetes, hypertension, and gestational age, play a vital role in LBW classification.

SHEL5K: An Extended Dataset and Benchmarking for Safety Helmet Detection.

Wearing a safety helmet is important in construction and manufacturing industrial activities to avoid unpleasant situations. This safety compliance can be ensured by developing an automatic helmet detection system using various computer vision and deep learning approaches. Developing a deep-learning-based helmet detection model usually requires an enormous amount of training data. However, there are very few public safety helmet datasets available in the literature, in which most of them are not entirely labeled, and the labeled one contains fewer classes. This paper presents the Safety HELmet dataset with 5K images (SHEL5K) dataset, an enhanced version of the SHD dataset. The proposed dataset consists of six completely labeled classes (helmet, head, head with helmet, person with helmet, person without helmet, and face). The proposed dataset was tested on multiple state-of-the-art object detection models, i.e., YOLOv3 (YOLOv3, YOLOv3-tiny, and YOLOv3-SPP), YOLOv4 (YOLOv4 and YOLOv4pacsp-x-mish), YOLOv5-P5 (YOLOv5s, YOLOv5m, and YOLOv5x), the Faster Region-based Convolutional Neural Network (Faster-RCNN) with the Inception V2 architecture, and YOLOR. The experimental results from the various models on the proposed dataset were compared and showed improvement in the mean Average Precision (mAP). The SHEL5K dataset had an advantage over other safety helmet datasets as it contains fewer images with better labels and more classes, making helmet detection more accurate.

A Wideband High-Isolation Microstrip MIMO Circularly-Polarized Antenna Based on Parasitic Elements.

This work presents a wideband, all-side square-cut square patch multiple-input, multiple-output circularly-polarized (MIMO-CP) high-isolation antenna. The MIMO-CP antenna contains a two-port square cut on all corners of the square patch, and parasitic elements of 9 × 5 periodic square metallic plates are designed and operated. The outer dimensions of the antenna are 40 × 70 mm2, and the FR4 substrate height is 1.6 mm. The proposed antenna with the parasitic elements improves impedance matching and enhances S-parameters and axial ratio (AR). In the suggested MIMO-CP antenna, a parasitic element is designed and placed around the antenna periodically to reduce mutual coupling (MC) and improve CP. Simulated results show that the suggested antenna has a wide bandwidth (BW) from 4.89 to 6.85 GHz for S11 and was < −10 dB with AR ≤ 3 dB from 5.42 to 6.58 GHz, with a peak gain of 6.6 dB. The suggested antennas have more than 30 dB isolation and a low profile, are affordable, easily made, and are CP. To make a comparison with the measured and simulated results, a MIMO-CP antenna structure was fabricated and tested. The suggested antenna is better in terms of efficiency, envelope correlation coefficient (ECC), diversity gain (DG), channel capacity loss (CCL), and total active reflection coefficient (TARC). The proposed antenna is adequate for WLAN applications.

High-Sensitivity Liquid Dielectric Characterization Differential Sensor by 1-Bit Coding DGS.

This paper presents two devices to detect the liquid dielectric characterization. The differential method was used to enhance the robustness and reduce tolerance. A basic sensor based on defected ground structure (DGS) was designed and the optimization for the squares of the DGS via adaptive genetic algorithm was applied to enhance the performance of the microwave sensor, which was shown by the difference of the two resonant frequencies. Furthermore, the electric field distribution was enhanced. Glass microcapillary tubes were used to hold samples to provide an environment of non-invasive. The optimized device exhibited the sensitivity of 0.076, which is more than 1.52 times than the basic structure. It could be considered a sensitive and robust sensor with quick response time for liquid dielectric characterization.

Performance Evaluation of Deep CNN-Based Crack Detection and Localization Techniques for Concrete Structures.

This paper proposes a customized convolutional neural network for crack detection in concrete structures. The proposed method is compared to four existing deep learning methods based on training data size, data heterogeneity, network complexity, and the number of epochs. The performance of the proposed convolutional neural network (CNN) model is evaluated and compared to pretrained networks, i.e., the VGG-16, VGG-19, ResNet-50, and Inception V3 models, on eight datasets of different sizes, created from two public datasets. For each model, the evaluation considered computational time, crack localization results, and classification measures, e.g., accuracy, precision, recall, and F1-score. Experimental results demonstrated that training data size and heterogeneity among data samples significantly affect model performance. All models demonstrated promising performance on a limited number of diverse training data; however, increasing the training data size and reducing diversity reduced generalization performance, and led to overfitting. The proposed customized CNN and VGG-16 models outperformed the other methods in terms of classification, localization, and computational time on a small amount of data, and the results indicate that these two models demonstrate superior crack detection and localization for concrete structures.

Wideband RCS Reduction Using Coding Diffusion Metasurface.

This paper presents a radar cross-section (RCS) reduction technique by using the coding diffusion metasurface, which is optimised through a random optimization algorithm. The design consists of two unit cells, which are elements '1' and '0'. The reflection phase between the two-unit cells has a 180° ± 37° phase difference. It has a working frequency band from 8.6 GHz to 22.5 GHz, with more than 9 dB RCS reduction. The monostatic RCS reduction has a wider bandwidth of coding diffusion metasurface as compared to the traditional chessboard metasurface. In addition, the bistatic performance of the designed metasurfaces is observed at 15.4 GHz, which shows obvious RCS reduction when compared to a metallic plate of the same size. The simulated and measured result shows the proficiency of the designed metasurface.