ML and DL Architectures Comparisons for the Classification of COVID-19 Using Chest X-Ray Images

Pandemic COVID-19 ranked as one of the world’s worst pandemics ever witnessed in history. It has affected every country by spreading this disease with an increase in mortality at alarming rates despite the technologically advanced era of medicine. AI/ML is one of the strong wings in the medical field so while fighting the battle to control and diagnose the best medicine for the outbreak COVID-19 disease. Automated and AI-based prediction models for COVID-19 are the main attraction for the scientist hoping to support some good medical decisions at this difficult time. However, mostly classical image processing methods have been implemented to detect COVID-19 cases resultant in low accuracy. In this chapter, multiple naïve machine and deep learning architectures are implied to evaluate the performance of the models for the classification of COVID-19 using a dataset comprising of chest x-ray images of, i.e., COVID-19 patients and normal (non-infected) individuals. The analysis looks at three machine learning architectures including Logistic Regression, Decision Tree (DT) Classifier, and support vector machine (SVM), and four deep learning architectures, namely: Convolutional neural networks (CNNs), VGG19, ResNet50, and AlexNet. The dataset has been divided into train, test and validation set and the same data have been used for the training, testing, and validation of all the architectures. The result analysis shows that AlexNet provides the best performance out of all the architectures. It can be seen that the AlexNet model achieved 98.05% accuracy (ACC), 97.40% recall, 98.03% F1-score, 98.68% precision, and 98.05% area under the curve (AUC) score. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Exploring Prediction of COVID-19 and its Severity using Machine Learning

COVID-19 has crippled the lives of millions in the world and is continuously doing so without any sight of relief. Even after the roll out of effective vaccines against COVID-19 and more than half of the population inoculated, it is still a widespread concern. This has led to extensive research around the world regarding the prediction of the COVID-19 disease, its diagnosis, developing drugs for its treatment and its forecasting, etc. Machine Learning has proved its significance in almost every domain and its techniques are also being actively used against COVID-19 by the researchers giving satisfactory results. In this paper, we have highlighted some of the efficient research that have been done using machine learning techniques to predict COVID-19 disease and its severity in patients. The performance of those techniques has been discussed and analyzed. We also carried out a comparative analysis of the most common techniques used in terms of accuracy obtained by them. It has been found that Support Vector Machines, Neural Networks and K-Nearest Neighbor models give the best performance in most of the research works. © 2022 IEEE.

Effect of teleworking on psychological health and employee performance during the COVID-19 pandemic

The aim of this study was to examine how the widespread use of telework during the COVID-19 pandemic may have created a specific work context influencing employees’ psychological health and performance. Results of analyses conducted on a sample of 3771 Canadian teleworkers revealed that telework created additional demands such as task interdependence and professional isolation. These demands had negative effects on telework performance by increasing the frequency of perceived stress. However, the presence of resources such as organizational support appeared to play a buffering role in moderating the direct effect of professional isolation on telecommuting performance. © 2022 AIPTLF

Patient experience of accessing a virtual cystic fibrosis service during the 2020 SARS-CoV-2 pandemic in Blackpool, North West of England

In 2015 Regional Specialist Commissioners awarded Blackpool Teaching Hospitals the commission of a new regional cystic fibrosis centre in the North West of England. The Blackpool Adult Cystic Fibrosis Service (BACFS) opened in February 2017 and was the first entirely new UK adult CF centre in three decades. Objective: To explore patient experiences of accessing a virtual adult cystic fibrosis service during the SARS-CoV-2 pandemic, to inform future practice and service development. Method: Between December 2020 and January 2021, 44 patients, accessing a blended approach to care through both virtual and face-to-face appointments, where necessary, were offered a CF-related adaptation of our hospital's Friends and Family Test. The questionnaire provided the mechanism to explore patient satisfaction and experience of BACFS during the pandemic (March 2020–December 2020) giving opportunity to critically appraise the service and their virtual experience. Results: Response rate was 18%. 100% of patients rated the service as either good or excellent, demonstrating care and compassion, being listened to, developing autonomy over their healthcare with shared action plans and improving confidence in the management of their condition. 50% of patients reported their virtual experience was ‘very good’ and 50% ‘good’, with 100% of patients reporting the frequency of their appointments was ‘just right’. Further qualitative themes of the virtual experience are presented, including convenience of appointments and feeling connected with the team during the pandemic. Conclusion: The data presented demonstrates that patients value the service provided by BACFS. Patients have been supportive of the virtual service provision with a suggestion that the offering may continue as we consider service delivery post SARS-CoV-2. The adapted patient experience questionnaire continues to drive service development and inform our approaches in alternative ways to engage our patients.

Embedding an electronic patient record into a developing UK service and impact during the COVID-19 pandemic - Blackpool Adult Cystic Fibrosis Service (BACFS)

Objectives: In December 2019 our new Service Manager successfully led BACFS’ transition from paper records to an Electronic Patient Record (EPR) called EMIS The clinical team had intermittent EPR engagement with Trust IT and Governance for 3 years prior with limited progress. This summarises the team's experience including impact during COVID-19. Methods: A questionnaire was sent to the BACFS multidisciplinary team (n = 13) asking for success scores (0–10) for 11 key areas and comments on challenges, lessons learnt and future development. Results: BACFS has successfully adopted EMIS and is now the principal recording system for CF clinical data, without which BACFS could not have worked remotely during COVID-19. The main challenges identified by users were IT authorisation/ Governance delays, lack of suitable IT, staff knowledge and use during inpatient episodes. Key lessons learned were it needs a designated project manager, a team ready for change, appropriate technology and a deadline. Recurrent user development suggestions included improvements to templates and inpatient processes. Conclusion: It has been a varied team experience yet hugely positive from a service perspective;EMIS has been pivotal for BACFS to function safely and effectively during COVID-19. None of the challenges were insurmountable with correct stakeholder engagement, investment in IT kit, peer support and, most importantly, an implementation lead. Challenges were easier to overcome in a service with low patient numbers and a small team, all of whom supported change. We are keen to further optimise our EPR use and share our experience with other services. [Table Presented]

Between the devil and the deep blue see: the role of situated optimism

The Covid-19 pandemic had unprecedented repercussions on many companies. The aim of this article is to study the influence of situated optimism on the entrepreneur's ability to overcome stress in times of crisis. Especially, when s/he is stuck between the devil (financial problems resulting from crisis measures) and the deep blue sea (financing obstacles). The research method adopted is a cross-sectional quantitative study on 677 SME managers and self-employed. The use of the partial least squares (PLS) structural equation method revealed the mediating effect of situated optimism on stress induced by financial problems. The financial issues encountered during the pandemic have been a source of stress for many entrepreneurs. The results show that in this context of crisis, situated optimism is a psychological resource that should not be neglected to enable them to cope with perceived stress.

Hot-Spot Zone Detection to Tackle Covid19 Spread by Fusing the Traditional Machine Learning and Deep Learning Approaches of Computer Vision

Corona Virus is a pandemic, and the whole world is affected due to it. Apart from the vaccine, the only cure for this drastic disease is to follow the rules and regulations that avoid further spread. There are different mechanisms like (Social Distancing, Mask Detection, Human occupancy etc.) through which we can able to stop the spread of the coronavirus. In this paper, we proposed hotspot zone detection using the computer vision techniques of deep learning. We have defined the hotspot area as the particular region on which the person touches more than some specified threshold. We further mark that area to some specific color to help the authority take necessary action and disinfect that particular place. To implement this algorithm, we have utilized the human-object interaction concept. We have extracted the dataset of person classes from the publicly available dataset for the person detection and the self-generated dataset to train the algorithm. Different experiments on object detection algorithms (YOLO-v3, Faster RCNN, SSD) for person detection have been performed in this work. We achieved the maximum accuracy in real-time on the YOLO-v3 for person detection. Whereas we have marked the specific area using the template matching algorithm of computer vision techniques. Our proposed algorithm detects the persons and extracts the region of interest points on which the user draws the rectangle. Then we find the intersection over union ratio between the detected person and the region of interest of the marked area to make the decision. We have achieved 88.72% accuracy on person detection in the local environment. Whereas, for the whole system of human-object interaction for detecting the hotspot area zone, we have achieved 86.7% accuracy using the confusion matrix.

What about the first public health crisis of 2019? EVALI in the Time of COVID

Introduction: E-cigarette or Vaping Product Use Associated Lung Injury (EVALI) and Coronavirus Disease 2019 (COVID-19) are both relatively new disease processes which can cause acute respiratory failure. This report describes the case of a 17-year-old male with a history of vaping cannabis during the first wave of the COVID-19 pandemic in Michigan. Report: A previously healthy 17-year-old male presented with cough, shortness of breath, chest pain, fever, and hypoxia requiring 40L high flow nasal cannula (HFNC) and 100% FiO2 to maintain oxyhemoglobin saturations of 88%. He showed no tachypnea or retractions, mimicking the “happy hypoxia” reported in COVID-19 patients. His physical exam demonstrated inspiratory crackles and migratory diminished breath sounds. Chest x-ray showed mild peri-bronchial thickening, subtle right perihilar opacities, and hyperexpansion. Bloodwork showed a procalcitonin of 10 nanograms/milliliter, ESR of 10 millimeters/hour, and CRP of 3.5 milligrams/deciliter. Three prior SARS-CoV2 tests were negative, but given clinical suspicion, the patient was treated as a Patient Under Investigation (PUI) for COVID-19 for 48 hours and re-tested. Care was aligned with institutional COVID-19 guidelines to minimize aerosol-generating procedures;diagnostic bronchoscopy, positive pressure ventilation, and transport for chest CT were discouraged, especially as our patient was awake, interactive, and with gradually improving trajectory. Infection prevention guidelines prohibited our patient's parents from leaving his room for a private interview, but friends alerted them to a history of vaping cannabis, which our patient corroborated. He was transitioned towards supportive care for EVALIinduced bronchoconstriction and improved with beta agonists, systemic steroids, and HFNC. After his fourth negative SARS CoV2 test, the patient underwent a high-resolution chest CT, which showed diffuse ground-glass opacities with subpleural sparing. He was discharged after four days with Pediatric Pulmonology follow up. He was counseled against further e-cigarette or cigarette use. Discussion: This case illustrates challenges in the diagnosis of EVALI during the COVID-19 pandemic, particularly among adolescents. Both conditions present with acute respiratory failure absent another source. Both can have significant hypoxia, elevated inflammatory markers, and an ARDS phenotype. Both demonstrate ground-glass opacities on CT scan. Importantly, here are differences in the workup, management, and public health implications of EVALI and COVID-19. Both are reportable to the Department of Public Health and warrant intervention: Anti-vaping campaigns and restriction on access for EVALI, infection control and immunization programs for COVID-19. As the world endeavors to contain the COVID-19 pandemic through surveillance, treatment, and immunization, we also hope to regain momentum against EVALI.