ABSTRACT
In recent years, due to the impact of COVID-19 around the world, there has been a serious shortage of medical resources. In order to supplement the manpower and fear that medical staff's contact with patients will cause a breach in the epidemic, reduce the workload of nurses, and help nurses perform repetitive tasks so that nurses can concentrate more on the patient's condition. Therefore, this paper proposes M-Robot, which is a friendly interface service robot based on the Android system and can be controlled by voice, touch, and remote control in the medical care field. The system is mainly divided into two parts. One is the web server. The web server is divided into two parts: front-end and back-end. The front-end is responsible for friendly user interface management, and the back-end is for accessing the SQLite database, as well as processing speech recognition and semantic understanding in voice services. In the other part, we use TEMI robot to develop and complete the desired service. Its service content includes environment introduction, delivery service, questionnaire survey, broadcast car, scheduling reminder, follow-up record, and patient instruction video. In the voice control mode, the user can say the wake-up word to the robot and say the required service content, and the robot will execute after receiving the message;in the remote control mode, we provide a friendly web interface for remote control. As well as the information needed to manage various services. © 2022 IEEE.
ABSTRACT
COVID-19 is highly contagious and highly pathogenic, It seriously threatens human life and health. Rapid detection of positive COVID-19 cases is very important in stopping the spread of the virus. At early diagnosis, It is the most simple and rapid indicator for judging changes in the illness. As the COVID-19 chest X-ray image dataset continues to expand, Researchers build a CNN-based COVID-19 detection model on Apache Spark. The model can effectively detect positive cases of COVID-19. This article first introduces the big data platform Apache Spark, Deep Learning Technology CNN, transfer learning techniques, etc. Then, it summarizes the characteristics and deficiencies of the research on chest X-ray image recognition of COVID-19 in recent years. Finally, Under the big data thinking, This paper proposes a technical direction for rapid detection of COVID-19 based on the big data analysis platform Apache Spark and the deep learning algorithm CNN for large-scale COVID-19 chest X-ray image datasets. © 2022 WCSE. All Rights Reserved.
ABSTRACT
Due to rapid change in influenza viruses, a prediction model for outbreaks of influenza-like illnesses helps to find out the spread of the illnesses in real time. In addition to using traditional hydrological and atmospheric data, popular search keywords on Google Trends are used as features in this research. Google Trends are popular keyword searches on the Google search engine. Popular keywords used in discussions of influenza-like symptoms at specific regions within specific periods are used in this research. Public holiday information in Taiwan, the population density, air quality indices, and the numbers of COVID-19 confirmed cases are also used as features in this research. An Ensemble Learning model, combining Random Forest and XGBoost, is used in this research. It can be confirmed from the actual experimental results in this research that the use of the ensemble learning prediction model proposed in this research can accurately predict the trend of influenza-like cases. The evaluation results show that the mean RMSLE of our proposed model is 0.2 in comparison with the actual number of influenza-like cases. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
ABSTRACT
The coronavirus disease 2019 (COVID-19) pandemic that started at the beginning of the year 2020 has significantly disrupted people’s daily life around the world. Understanding and quantifying the impact of such a large-scale disruption will help people mitigate the pandemic and enhance the resilience for future preparation of similar events. In this chapter, we present a research work studying the impact of COVID-19 on public transport in terms of bus delay, which involves big data processing and analysis on multisource datasets containing COVID-19 case data, bus GTFS (General Transit Feed Specification) data, and LGA (Local Government Area) boundary data. The datasets in use are heterogeneous, arrive in large volumes and in real time, and have a spatiotemporal distribution, which brings true challenges to this research. To quantify the bus delay changes, we propose a methodology consisting of real-time data crawling, map-matching, arrival time estimation, and bus delay calculation and aggregation. The methodology is applied to a case study focusing on the Sydney metropolitan region across different stages of the COVID-19 pandemic from February to March 2020. The case study shows that during March 2020, the COVID-19 pandemic has significantly impacted people’s travel behaviors in Sydney, but the influence varies in different areas. The most affected areas are the central and eastern suburbs, which recorded a drop of 9.5 min of bus delay during afternoon peak hours. The findings are helpful to understand and mitigate the restriction impact in different city areas with different conditions. The quantified delay reduction also reveals the potential of better transport performance, which could be used as a benchmark of transport performance improvement after the pandemic. The main contributions of this work include the methodology to quantify travel behavior changes under large disruptions such as COVID-19 pandemic and the case study on large-scale and long-period travel behavior shift that seldom happened before. © 2022 Elsevier Inc.
ABSTRACT
Purpose - When the coronavirus disease 2019 (COVID-19) pandemic seriously hit the USA, a lot of cities/states announced their lockdowns, in some cases forbidding employees to go to work. But workers in the so called "essential sectors" were exempt from the order, and on the contrary were required to remain on the job in order to maintain the services and functions considered vital to the community. If they have not been paid well in comparison to those in the other sectors, there would be a stronger case for granting them a special hazard pay during the pandemic. This paper aims to design a way to measure the "importance" or being "essential" of the different sectors in the economy, and then investigates whether the actual pay of the workers in these sectors is consistent with the measured importance. Design/methodology/approach - At least two policy issues emerged from such an arrangement: (1) How can one define the "essential sectors" objectively instead of the authorities preparing a list according to their administrative procedure? (2) How well have been the workers in the essential sectors paid before the pandemic strike? The concept of a revised Leontief forward linkage effect will be used in an input-output model to gauge the relative "importance" of the different sectors in the US economy. Then the measured importance will be compared with the average compensation of the employees in these sectors. Findings - It is found that for some sectors such as agriculture, retail trade, and repair and installation of machinery and equipment the ratio of workers' compensation relative to the national average is substantially lower than the relative importance of the sectors employing them. That is, many of them have been substantially underpaid in spite of their importance. Research limitations/implications - The scope of this study is limited to one country, the USA, but the methodology can be applied to other countries as well. Originality/value - This study is an original research that contributes to an improved understanding of the importance of the workers engaged in different sectors in the USA during COVID-19.
ABSTRACT
Purpose : Virtual reality-based oculokinetic perimetry (VR-OKP) is a mobile, short, screening visual field test. The purpose of this study is to examine whether VR-OKP could be implemented in a remote fashion and used to assess the stability of the visual field. Methods : Glaucoma subjects with known glaucomatous defects who had taken an inperson VR-OKP test in 2019 were re-enrolled and underwent the same test remotely in 2020. An exit survey comparing patients' preferences for visual field testing was also conducted remotely. Subjects underwent Humphrey visual field 24-2 (HVF) testing within 3 months of VR-OKP both in 2019 and 2020. For unadjusted comparisons between HVF and VR-OKP, we compared mean sensitivities and fraction of points. A non-parametric bootstrap analysis that resampled eyes with replacement was done to calculate the 95% confidence interval of the Spearman correlation coefficient between HVF sensitivity versus VR-OKP fraction seen at each point of the 54 test locations. Results : The cohort consisted of 19 eyes of 11 patients (55% female, 73% Caucasian, 27% Asian, mean age 61.4 ± 12.6 yrs) with moderate to advanced glaucoma (2020 average HVF mean deviation -4.23 dB ± 5.12). VR-OKP from 2019 to 2020 had a decreased mean percent change of -6.31% ± 17.22 (p=0.13) compared to HVF testing from 2019 to 2020 which had an increased mean change of mean deviation of +2.41 dB ± 1.35 (p<.00005). The Pearson's Correlation Coefficient between 2019 VR-OKP fraction seen and 2019 HVF mean sensitivity was 0.74, while it was 0.56 in 2020. Spearman correlation coefficients of HVF sensitivity vs VR-OKP fraction seen at each point ranged from -0.01-0.86 (median = 0.33). Subjects found VR-OKP to be as comfortable as HVF (p=0.8) and less fatiguing (p=0.03). Conclusions : This study highlights the feasibility of a remote option for visual field assessment. The correlation between VR-OKP and HVF in 2019 was higher than in 2020. One explanation is that the 2019 test was taken with in-person instruction, whereas the 2020 test was administered remotely. While the mean percentage change in VR-OKP was non-significant, the 2019 to 2020 change in HVF was statistically significant and showed improvement, which may be due to long term fluctuation. This short VR-OKP test is less fatiguing to patients, can detect non-progression even when taken at home, and could potentially be deployed to decrease patients' risk of COVID-19.
ABSTRACT
Purpose : To 1) determine the feasibility of remotely administering and training subjects at home on how to use Vivid Vision Perimetry (VVP-10), a portable virtual reality-based visual field test during COVID-19 shelter-in-place;2) assess the correlation between VVP-10 and standard automated perimetry (SAP) and the test-retest variability of VVP-10. Methods : Inclusion criteria included subjects 21 or older with glaucomatous visual field defects, and exclusion criteria included those with retinal diseases or significant cataracts. Virtual reality devices were given to subjects during clinic visits or mailed to them. Subjects were remotely trained using training software and coaching via Zoom and proceeded to take 10 tests at home over 14 days. Subject age and sex, test duration, response rate at each of the test points (fraction correct), and SAP results including mean deviation and individual test point sensitivities were recorded. The Pearson correlation coefficients of SAP mean sensitivity versus VVP-10 fraction correct for all eyes together and fraction correct of tests 6-10 versus tests 1-5 for individual eyes were calculated. A bootstrap analysis that resampled eyes with replacement was done to calculate the 95% confidence interval of the Spearman correlation coefficient between SAP sensitivity versus VVP-10 fraction correct for each point of the 54 test locations. Results : Of the 20 subjects enrolled, 11 (55%) were male and the average (SD) age was 62.9 (10.5) years. Eight (40%) subjects were Asian and 12 (60%) were Caucasian. In total, 37 glaucomatous eyes with an average (SD) mean deviation of-6.1 (6.1) dB were analyzed. The Pearson correlation coefficient of SAP mean sensitivity versus VVP-10 fraction correct for 37 eyes was 0.68. The Pearson correlation coefficient between fraction correct in tests 6-10 versus tests 1-5 for individual eyes ranged from 0.78-0.99 (median = 0.94). Spearman correlation coefficients of SAP sensitivity versus VVP-10 fraction correct at each point ranged from 0.008-0.85 (median = 0.58). Conclusions : We demonstrate good test-retest variability of VVP-10 and a strong correlation with SAP, both globally and in a pointwise manner. VVP-10 is portable, inexpensive, and can be used entirely remotely while producing results that are similar to the current gold standard for assessing glaucomatous visual fields.