Motorola (South Korea)

Timely detection of ventricular fibrillation (VF) is essential in the fight for deadly cardiac arrhythmias. This research introduces a wearable electrocardiogram (ECG)-based VF diagnosis system using Convolutional Neural Networks (CNNs) for real-time evaluation. The proposed CNN model autonomously derives spatial and temporal patterns from raw ECG data, hence improving detection accuracy, in contrast to previous approaches that depend on created features. The model was trained and assessed using a publically accessible ECG dataset, attaining an accuracy of 98.5%, sensitivity of 97.8%, and specificity of 99.1% in differentiating VF from normal and other arrhythmic states. Data augmentation methods and refined hyperparameters enhanced generalization across various ECG signal variances. Further validation was conducted on the system utilising wearable ECG data collected in real-time, with an average inference time of 12 ms per sample, enabling quick identification appropriate for continuous monitoring applications. The proposed CNN-based method markedly improves traditional rule-based and machine-learning methods regarding performance and computing efficiency. Incorporating this model into wearable health monitoring devices may enable early diagnosis of VF, decrease reaction time for medical treatment, and improve the results for patients. Future efforts will focus on practical implementation and improving power efficiency for edge computing.

Physical activity during pregnancy is established to be a safe and effective strategy to improve maternal and neonatal health. The objective of this article was to investigate the effects of physical activity on the mother, foetus and delivery outcomes. We reviewed literature on physiological adaptations to exercise during pregnancy, maternal and foetal outcomes, and the influence of physical activity on labour duration, mode of delivery, oxytocin administration, perineal trauma, and other perinatal parameters. The findings suggest that regular, moderate physical activity may reduce the risk of gestational complications such as diabetes and hypertension, shorten labour duration, decrease the risk of operative delivery, reduce the need for analgesia and have a beneficial effect on the foetus. Further randomized controlled trials are required to confirm the association between physical activity and pregnancy outcomes, possibly including the differentiation between exercises of various intensity, as well as to provide standardized recommendations.

Globallashib borayotgan dunyoda reklama bilan bog'liq masalalar muhimroq bo'lib bormoqda. Reklamaning kompaniya mahsulotlarini targ'ib qilishdagi samaradorligi keng tarqalishiga olib keldi, butun dunyo bo'ylab reklama va bu reklamaning katta qismi ingliz tilidan foydalanadi. Bu turli xil deb hisoblangan reklama ingliz tilining paydo bo'lishiga yo'l ochadi. Reklama uchun ingliz tili juda o'ziga xos xususiyatlarga ega. Lingvistik va diskursiv xususiyatlar reklama ingliz tilining o'ziga maxsus nutq xos janri ekanligini ko'rsatadi. Shunday qilib, ushbu tadqiqotning maqsadi reklama tilini tahlil qilishdir. Reklamaning turli lingvistik va diskursiv xususiyatlarini ingliz uchun aniqlaydi. Shu maqsadda joriy tadqiqot ingliz tilidagi reklama matnlari namunasini tanlab oldi. Ushbu maqola uchun ingliz tilining turli xil asosiy lingvistik va diskursiv xususiyatlarini taqdim etish reklama bilan yakunlanadi.

This paper uses machine learning to predict either the voltage, current, or state of charge (SOC) during a discharging of a Li-Ion cell. Due to the non-linear characteristics of a Li-Ion cells, machine learning is used to create a model for prediction. Predicting the Li-Ion cell characteristics is useful for products in determining the end of discharge (EOD) levels under different loading conditions. The model used is a Deep Neural Network (DNN) with 2 hidden layers with 128 nodes each and is implemented using Python. To train the model, approximately 6000 data points of charging and discharging data of a LCO prismatic CS2 1100mAh Li-Ion cell is used. It was found that the accuracy of the model is approximately 5% and worsens at lower SOC.

Reconfigurable intelligent surfaces (RISs) have attracted increasing interest due to their ability to improve the coverage, reliability, and energy efficiency of millimeter wave (mmWave) communication systems. However, designing the RIS beamforming typically requires large channel estimation or beam training overhead, which degrades the efficiency of these systems. In this paper, we propose to equip the RIS surfaces with visual sensors (cameras) that obtain sensing information about the surroundings and user/basestation locations, guide the RIS beam selection, and reduce the beam training overhead. We develop a machine learning (ML) framework that leverages this visual sensing information to efficiently select the optimal RIS reflection beams that reflect the signals between the basestation and mobile users. To evaluate the developed approach, we build a high-fidelity synthetic dataset that comprises co-existing wireless and visual data. Based on this dataset, the results show that the proposed vision-aided machine learning solution can accurately predict the RIS beams and achieve near-optimal achievable rate while significantly reducing the beam training overhead.

This paper is part of the ACP2023 Conference Proceedings (View) Full Paper View / Download the full paper in a new tab/window

Access to the Internet is necessary to ensure equity for the right to education. However, about 75% of school-age children in rural areas of the world do not have access to the Internet at home. In the Ecuadorian context, only 5.1% of the rural population uses the Internet in public/state schools. Therefore, actions are necessary to reduce the digital gap to improve the quality of education in rural communities in Ecuador. An initiative of IEEE Ecuador, with the support of IEEE SIGHT and the Municipality of Nabón, allowed the rehabilitation of a data network in 40 schools in the Nabón community in 2020. This paper examines the impact of the project, through of the collection of quantitative and qualitative data through stakeholder surveys, to evaluate the impact of data network rehabilitation in the community during the first year of operation. The results show that the Internet in schools has improved the quality of education for students, has allowed children and young people to have access to online education during the COVID-19 pandemic, and has decreased the dropout rate in the benefited schools. The Internet service in schools has also allowed families to save money and entertain themselves. The paper also reports on the evaluation of the social impact of the project, through a Social Return On Investment (SROI) analysis.

Resistive-type pressure sensor, which are mainly utilized in industry, are easy to manufacture and are not significantly affected by external electromagnetic fields, unlike capacitive type. However, the produce signal is not linear, and it is also difficult to measure a wide range of pressures using such a sensor. Therefore, before being utilized, the extracted nonlinear data from them need to be processed by. A resistive sensor that is capable of measuring a wide range of pressure of up to 4 MPa with constant linearity is presented. Moreover it can selectively control the sensing pressure band, or act as an on/off switch, without the need for any additional computer processing.

SummaryIn this paper, development of a small signal model for 2 × 200 μm GaN HEMT based on the conventional 20‐element model is presented. The proposed model presents a direct parameter extraction algorithm, instead of the hybrid optimization approach, that provides simplification, accuracy, and less computational complexity. The extrinsic elements are extracted using a modified cold pinch‐off condition while discarding the unwanted forward biasing of the gate. The negative drain to source capacitance Cds is also observed in the ohmic region (for smaller VDS). An excellent agreement found between the measured and modeled data for a wide range of frequencies and bias values shows the effectiveness of the proposed approach. The proposed modeling technique is validated with a good agreement between the achieved bias dependency of intrinsic parameter values and the respective theoretical parameter values.