Your browser doesn't support javascript.
Show: 20 | 50 | 100
Results 1 - 20 de 2.201
Filter
Add filters

Year range
1.
3rd International Workshop on Experience with SQuaRE Series and Its Future Direction, IWESQ 2021 ; 3114:23-28, 2021.
Article in English | Scopus | ID: covidwho-1824439

ABSTRACT

In a context where the availability of information represents the opportunity for companies to gain a competitive advantage in the market through the use of sophisticated AI algorithms, data quality assumes a strategic role. With this paper we want to show that the adoption of an international quality measurement standard such as the one present in the SQuaRE series can on the one hand improve the ethical aspect of machine learning algorithms and on the other hand meet the requirements imposed by the European Community regarding the protection of personal data of citizens in Member States (GDPR). Indeed, although the attention to the protection of personal data is mainly directed towards the aspects of security and confidentiality, in a holistic view we should also evaluate the risks arising from the absence of quality in the data. In this context, we consider consistent and of reference for the international community the choice of the Italian legislator made for the Public Administrations. Since 2013 the Agency for Digital Italy (AgID) has suggested the adoption of ISO/IEC 25012 for public administrations in charge of managing databases of national interest. In the article, we propose a methodological approach that ensures the governance of data quality and some open questions regarding the homogeneity of the selected measures. © 2021 for this paper by its authors

2.
RISTI - Revista Iberica de Sistemas e Tecnologias de Informacao ; 2021(E45):200-211, 2021.
Article in Spanish | Scopus | ID: covidwho-1823818

ABSTRACT

Twitter is an important social network and information channel where opinions (tweets) can be obtained and processed in real time that can be explored, analyzed and organized to make better decisions. Opinion mining is a natural language processing task that identifies user opinions as positive, negative, or neutral. COVID-19 is an infectious disease caused by the coronavirus that appeared in December 2019 in China and immediately provoked a large number of opinions. To allow Panamanian health organizations to detect opportunities to improve the quality of medical care, we propose to classify the tweets the analysis of two approaches: deep learning and machine learning for to appreciate which is more precise. We obtained encouraging results with a precision of 95.6%. © 2021, Associacao Iberica de Sistemas e Tecnologias de Informacao. All rights reserved.

3.
Frontiers in Cellular and Infection Microbiology ; 12, 2022.
Article in English | EMBASE | ID: covidwho-1822355

ABSTRACT

The Coronavirus Disease 2019 (COVID-19) has spread all over the world and impacted many people’s lives. The characteristics of COVID-19 and other types of pneumonia have both similarities and differences, which confused doctors initially to separate and understand them. Here we presented a retrospective analysis for both COVID-19 and other types of pneumonia by combining the COVID-19 clinical data, eICU and MIMIC-III databases. Machine learning models, including logistic regression, random forest, XGBoost and deep learning neural networks, were developed to predict the severity of COVID-19 infections as well as the mortality of pneumonia patients in intensive care units (ICU). Statistical analysis and feature interpretation, including the analysis of two-level attention mechanisms on both temporal and non-temporal features, were utilized to understand the associations between different clinical variables and disease outcomes. For the COVID-19 data, the XGBoost model obtained the best performance on the test set (AUROC = 1.000 and AUPRC = 0.833). On the MIMIC-III and eICU pneumonia datasets, our deep learning model (Bi-LSTM_Attn) was able to identify clinical variables associated with death of pneumonia patients (AUROC = 0.924 and AUPRC = 0.802 for 24-hour observation window and 12-hour prediction window). The results highlighted clinical indicators, such as the lymphocyte counts, that may help the doctors to predict the disease progression and outcomes for both COVID-19 and other types of pneumonia.

4.
18th IEEE International Symposium on Biomedical Imaging (ISBI) ; : 155-159, 2021.
Article in English | Web of Science | ID: covidwho-1822030

ABSTRACT

The outbreak of COVID-19 has led to a global effort to decelerate the pandemic spread. For this purpose chest computed-tomography (CT) based screening and diagnosis of COVID-19 suspected patients is utilized, either as a support or replacement to reverse transcription-polymerase chain reaction (RT-PCR) test. In this paper, we propose a fully automated AI based system that takes as input chest CT scans and triages COVID-19 cases. More specifically, we produce multiple descriptive features, including lung and infections statistics, texture, shape and location, to train a machine learning based classifier that distinguishes between COVID-19 and other lung abnormalities (including community acquired pneumonia). We evaluated our system on a dataset of 2191 CT cases and demonstrated a robust solution with 90.8% sensitivity at 85.4% specificity with 94.0% ROC-AUC. In addition, we present an elaborated feature analysis and ablation study to explore the importance of each feature.

5.
11th International Conference on Indoor Positioning and Indoor Navigation (IPIN) ; 2021.
Article in English | Web of Science | ID: covidwho-1822026

ABSTRACT

Smartphone apps for exposure notification and contact tracing have been shown to be effective in controlling the COVID-19 pandemic. However, Bluetooth Low Energy tokens similar to those broadcast by existing apps can still be picked up far away from the transmitting device. In this paper, we present a new class of methods for detecting whether or not two WiFi-enabled devices are in immediate physical proximity, i.e. 2 or fewer meters apart, as established by the U.S. Centers for Disease Control and Prevention (CDC). Our goal is to enhance the accuracy of smartphone-based exposure notification and contact tracing systems. We present a set of binary machine learning classifiers that take as input pairs of Wi-Fi RSSI fingerprints. We empirically verify that a single classifier cannot generalize well to a range of different environments with vastly different numbers of detectable Wi-Fi Access Points (APs). However, specialized classifiers, tailored to situations where the number of detectable APs falls within a certain range, are able to detect immediate physical proximity significantly more accurately. As such, we design three classifiers for situations with low, medium, and high numbers of detectable APs. These classifiers distinguish between pairs of RSSI fingerprints recorded 2 or fewer meters apart and pairs recorded further apart but still in Bluetooth range. We characterize their balanced accuracy for this task to be between 66.8% and 77.8%.

6.
6th International Conference on Advances in Biomedical Engineering (ICABME) ; : 222-227, 2021.
Article in English | Web of Science | ID: covidwho-1822025

ABSTRACT

In the light of the rapidly growing COVID-19 pandemic, the need for an expeditious diagnosis of COVID-19 infection became essential. The immediate diagnosis will allow the initiation of the isolation process and adequate treatment as well. While the standard test used for the diagnosis of COVID19 disease (RT-PCR) is usually time consuming (6 hours up to days in some centers);the need for a highly sensitive test became essential. Many studies have illustrated the utility of chest CT scan in the diagnoses of COVID-19. This paper evaluates the value of classical machine learning techniques and the convolutional neural networks in aiding physicians to further classify patients into either COVID-19 positive or negative according to their chest CT findings, and thus facilitating their work. To address this problem, this paper proposes classical neural networks using statistical features and deep CNN models to further classify a dataset of preprocessed chest CT images, using several classifiers and to evaluate the results. This latter showed that the best proposed method was a four layers CNN with SVM classifier with 99.6% accuracy. This demonstrates the potential of the proposed technique in computer-aided diagnosis for healthcare applications, especially for COVID-19 classification.

7.
6th International Conference on Advances in Biomedical Engineering (ICABME) ; : 213-218, 2021.
Article in English | Web of Science | ID: covidwho-1822018

ABSTRACT

The current global spread of COVID-19, a highly contagious disease, has challenged healthcare systems and placed immense burdens on medical staff globally. Almost 5% to 10% among hospitalized patients will require ICU admission. Predicting ICU admission can help in managing better the patient and the healthcare system. This study aims to develop a model that can predict whether a COVID-19 patient, who has already been admitted to the hospital, will enter the ICU or not. This could be accomplished by monitoring his vital signs, and blood tests, and inquiring about his demographic records, during his stay in the hospital. Multiple models, including Artificial Neural Networks, Logistic Regression, Decision Tree, Random Forest, Gaussian Naive Bayes, Gradient Boosting, and Support Vector Machines, were designed and implemented using MATLAB and Python. Random Forest, Decision Tree, and Gradient Boosting, are examples of decision tree-based algorithms that outperformed the others. The Random Forest (Accuracy: 99.12%, Cross-Validation Accuracy 86.34%) and Decision Tree (Accuracy: 99.12%, Cross-Validation Accuracy 79.48%) and Gradient Boosting (Accuracy: 93.77%, Cross-Validation Accuracy: 86.96%) had the highest accuracy scores as compared to other models such as the Support Vector Machines (Accuracy: 87.74%, Cross-Validation Accuracy 72.42%). In future work, the aim will be to predict whether a patient will join ICU or not, based on monitoring for multiple windows. As a result, high accuracy scores will be reached, since the model will analyze the vital signs and laboratory data at multiple stages and timings. In this way, anticipating the requirement for ICU admission well ahead of time.

8.
JACC: Advances ; : 100043, 2022.
Article in English | ScienceDirect | ID: covidwho-1821317

ABSTRACT

Background COVID-19 infection carry significant morbidity and mortality. Current risk prediction for complications in COVID-19 is limited and existing approaches fail to account for the dynamic course of the disease. Objectives To develop and validate the COVID-HEART predictor, a novel continuously-updating risk prediction technology to forecast adverse events in hospitalized patients with COVID-19. Methods Retrospective registry data from patients with SARS-CoV-2 infection admitted to 5 hospitals were used to train COVID-HEART to predict all-cause mortality/cardiac arrest (AM/CA) and imaging-confirmed thromboembolic events (TE) (2550 and 1854 patients, respectively). To assess COVID-HEART’s performance in the face of rapidly changing clinical treatment guidelines, an additional 1100 and 796 patients, admitted after the completion of development data collection, were used for testing. Leave-hospital-out validation was performed. Results Over 20 iterations of temporally-divided testing, the mean AUROCs were 0.917 (95% CI: 0.916-0.919) and 0.757 (95% CI: 0.751-0.763) for prediction of AM/CA and TE, respectively. The interquartile ranges of median early warning times were 14-21 hours for AM/CA and 12-60 hours for TE. The mean AUROCs for the left-out hospitals were 0.956 (95% CI: 0.936-0.976) and 0.781 (95% CI: 0.642-0.919) for prediction of AM/CA and TE, respectively. Conclusions The continuously updating, fully interpretable COVID-HEART predictor accurately predicts AM/CA and TE within multiple time windows in hospitalized COVID-19 patients. In its current implementation the predictor can facilitate practical, meaningful change in patient triage and resource allocation by providing real-time risk scores for these outcomes. The potential utility of the predictor extends to COVID-19 patients post-hospitalization and beyond COVID-19.

9.
Computers in Biology and Medicine ; : 105587, 2022.
Article in English | ScienceDirect | ID: covidwho-1821197

ABSTRACT

Recent years have seen deep neural networks (DNN) gain widespread acceptance for a range of computer vision tasks that include medical imaging. Motivated by their performance, multiple studies have focused on designing deep convolutional neural network architectures tailored to detect COVID-19 cases from chest computerized tomography (CT) images. However, a fundamental challenge of DNN models is their inability to explain the reasoning for a diagnosis. Explainability is essential for medical diagnosis, where understanding the reason for a decision is as important as the decision itself. A variety of algorithms have been proposed that generate explanations and strive to enhance users' trust in DNN models. Yet, the influence of the generated machine learning explanations on clinicians' trust for complex decision tasks in healthcare has not been understood. This study evaluates the quality of explanations generated for a deep learning model that detects COVID-19 based on CT images and examines the influence of the quality of these explanations on clinicians’ trust. First, we collect radiologist-annotated explanations of the CT images for the diagnosis of COVID-19 to create the ground truth. We then compare ground truth explanations with machine learning explanations. Our evaluation shows that the explanations produced. by different algorithms were often correct (high precision) when compared to the radiologist annotated ground truth but a significant number of explanations were missed (significantly lower recall). We further conduct a controlled experiment to study the influence of machine learning explanations on clinicians' trust for the diagnosis of COVID-19. Our findings show that while the clinicians’ trust in automated diagnosis increases with the explanations, their reliance on the diagnosis reduces as clinicians are less likely to rely on algorithms that are not close to human judgement. Clinicians want higher recall of the explanations for a better understanding of an automated diagnosis system.

10.
Computers ; 11(4):18, 2022.
Article in English | Web of Science | ID: covidwho-1820191

ABSTRACT

During the COVID-19 epidemic, Twitter has become a vital platform for people to express their impressions and feelings towards the COVID-19 epidemic. There is an unavoidable need to examine various patterns on social media platforms in order to reduce public anxiety and misconceptions. Based on this study, various public service messages can be disseminated, and necessary steps can be taken to manage the scourge. There has already been a lot of work conducted in several languages, but little has been conducted on Arabic tweets. The primary goal of this study is to analyze Arabic tweets about COVID-19 and extract people's impressions of different locations. This analysis will provide some insights into understanding public mood variation on Twitter, which could be useful for governments to identify the effect of COVID-19 over space and make decisions based on that understanding. To achieve that, two strategies are used to analyze people's impressions from Twitter: machine learning approach and the deep learning approach. To conduct this study, we scraped Arabic tweets up with 12,000 tweets that were manually labeled and classify them as positive, neutral or negative feelings. Specialising in Saudi Arabia, the collected dataset consists of 2174 positive tweets and 2879 negative tweets. First, TF-IDF feature vectors are used for feature representation. Then, several models are implemented to identify people's impression over time using Twitter Geo-tag information. Finally, Geographic Information Systems (GIS) are used to map the spatial distribution of people's emotions and impressions. Experimental results show that SVC outperforms other methods in terms of performance and accuracy.

11.
Algorithms ; 15(4):13, 2022.
Article in English | Web of Science | ID: covidwho-1820152

ABSTRACT

In recent years, the topic of contactless biometric identification has gained considerable traction due to the COVID-19 pandemic. One of the most well-known identification technologies is iris recognition. Determining the classification threshold for large datasets of iris images remains challenging. To solve this issue, it is essential to extract more discriminatory features from iris images. Choosing the appropriate loss function to enhance discrimination power is one of the most significant factors in deep learning networks. This paper proposes a novel iris identification framework that integrates the light-weight MobileNet architecture with customized ArcFace and Triplet loss functions. By combining two loss functions, it is possible to improve the compactness within a class and the discrepancies between classes. To reduce the amount of preprocessing, the normalization step is omitted and segmented iris images are used directly. In contrast to the original SoftMax loss, the EER for the combined loss from ArcFace and Triplet is decreased from 1.11% to 0.45%, and the TPR is increased from 99.77% to 100%. In CASIA-Iris-Thousand, EER decreased from 4.8% to 1.87%, while TPR improved from 97.42% to 99.66%. Experiments have demonstrated that the proposed approach with customized loss using ArcFace and Triplet can significantly improve state-of-the-art and achieve outstanding results.

12.
28th IEEE International Conference on Electronics, Circuits, and Systems (IEEE ICECS) ; 2021.
Article in English | Web of Science | ID: covidwho-1819833

ABSTRACT

Coronaviruses are a large viral family that attacks key organs, particularly the lungs. The infection spread is growing by the day, affecting almost every industry. Various Artificial Intelligence studies have been proposed, to learn the measurable information of people who have been affected with COVID-19 and those who have recovered, as well as the death rate. Various data samples like chest images, lung images, swab results, blood samples, and CT scans are used to predict the COVID-19. The paper gives an in-depth look at how AI and machine learning techniques can be used to accurately predict COVID-19. The proposed review is centered around investigating the different AI methods, models, and logical registering procedures used in foreseeing the COVID-19 sickness. The study also summarizes the difficulties associated with current methods and future exploration works.

13.
International Conference on Decision Aid Sciences and Application (DASA) ; 2021.
Article in English | Web of Science | ID: covidwho-1819811

ABSTRACT

The novel coronavirus is the most crucial pandemic that has been faced in recent times. The pandemic has caused severe economic and social devastation, and due to the lack of experience, many countries have a huge burden to protect their people from the coronavirus. This virus spreads from person to person so easily via the droplets. Hence, it is hard to identify the virus if they do not show the symptoms;thus, the asymptomatic person can become a super spreader and spread the virus faster than a symptomatic person. It is essential to seek a technological-based application that can be globally used. Therefore, this paper proposes a machine learning-based speech signal processing application to identify the asymptomatic and mild symptomatic COVID-19 virus-infected individuals. The proposed method uses gammatone cepstral features along with eight machine learning methods for the classification task. Finally, the best-trained machine learning model will apply in a real-time-based speech signal processing application. The final results show that the K-Nearest Neighbour (KNN) and Ensemble Bagged Tree (EBT) can provide better classification results than the other machine learning models.

14.
International Conference on Decision Aid Sciences and Application (DASA) ; 2021.
Article in English | Web of Science | ID: covidwho-1819807

ABSTRACT

In this paper, we propose a performance assessment of a forest induction method called IDTNIM-RF which uses the IDT_NIM "Induction of Decision Tree New Information Measure" for the trees induction. The specific techniques applied to forests such as bagging and random feature selection are used for generating multiple IDT_NIM trees. We compare the IDTNIM-RF method with a RF random forest that uses CART as a basic rule. The success rate criterion of the single CART and IDT_NIM trees used respectively to generate URF and IDTNIM-RF forest sets is used as a performance assessment basis. To achieve this evaluation, data sets are carried from the UCI Repository and some learning bases we have already developed such MONITDIAB and COVID_EHU.

15.
International Conference on Decision Aid Sciences and Application (DASA) ; 2021.
Article in English | Web of Science | ID: covidwho-1819804

ABSTRACT

The coronavirus is a contagious disease and can spread very rapidly if the proper measures are not taken. Though the invention of the vaccines against coronavirus has given a sigh of relief, however, the complete eradication still looks a very long way to go. With the presence of the new variants of the coronavirus, the risk of the spread still remains. Among several guidelines given by the WHO and healthcare practitioners, facemasks have been one of the most effective ways to prevent the spread of the virus. However, some people usually ignore or forget to follow these guidelines especially in public places such as offices, shopping malls, etc. The number of people in such places is usually high and facemask is a factor to consider against the spread of the virus. Therefore, to hinder the spread of the virus, people with no facemask must be identified and notified. This research proposes a convolutional neural network-based deep learning model for detecting the people without facemasks using the frames captured from the livestream surveillance video. The research primarily focuses on the facemask detection module of the proposed system. The data for this study contains almost 1500 images for masked and without mask faces. The proposed model has been implemented using two different optimizers. The RMSprop optimizer-based model outperforms the Adam optimizer-based model. The accuracy achieved by RMSprop based model was 92.27% and the accuracy achieved by Adam optimizer-based model was 85.1%.

16.
Sustainable Cities and Society ; : 103929, 2022.
Article in English | ScienceDirect | ID: covidwho-1819603

ABSTRACT

To simultaneously promote health, economic, and environmental sustainability, a number of cities worldwide have established bike-sharing systems (BSS) that complement the conventional public transport systems. As the rapid spread of COVID-19 becoming a global pandemic disrupted urban mobility due to government-imposed lockdowns and the heightened fear of infection in crowded spaces, populations were increasingly less likely to use public transportation and instead shifted toward alternative transport systems, including BSS. In this study, we use probabilistic machine learning in a quasi-experimental research design to identify how the relevance of a comprehensive set of factors to predict the use of Bicing (the BSS in Barcelona) may have changed as COVID-19 unfolded. We unpack the key factors in predicting the use of Bicing, uncovering evidence of increasing bike-related built infrastructure (e.g., tactical urbanism), trip distance, and the income levels of neighborhoods as the most relevant predictors. Moreover, we find that the relevance of the factors in predicting Bicing usage has generally decreased during the global pandemic, suggesting altered societal behavior. Our study enhances the understanding of BSS and societal behavior, which can have important implications for developing resilient programs for cities to adopt sustainable practices through transport policy, infrastructure planning, and urban development.

17.
5th International Conference on Computational Vision and Bio Inspired Computing (ICCVBIC) ; 1420:367-380, 2021.
Article in English | Web of Science | ID: covidwho-1819415

ABSTRACT

Air pollution imposes a big environmental and health hazard challenge in the world. This necessitates a tool to be developed that can measure and analyze air quality in real time. This is also essential in quick restoration and assessment of lungs during the prevailing COVID pandemics. The real-time air quality indexes of gases like carbon monoxide, ozone, SO2, NO2, etc., are collected and sent to the cloud (firebase). The collected data is analyzed using machine learning algorithms which leads to the prediction of severity of air quality contamination. Further in this paper, we propose an analysis of location wise air pollution measurement and accuracy in the prediction of air quality.

18.
5th International Conference on Computational Vision and Bio Inspired Computing (ICCVBIC) ; 1420:239-251, 2021.
Article in English | Web of Science | ID: covidwho-1819414

ABSTRACT

Coronavirus disease 2019 (COVID-19) is an infectious disease caused by severe acute respiratory syndrome coronavirus2. COVID-19 has created the worldwide pandemic situation and it is causing a greater health crisis and deaths of the millions of humans all over the world. All the socio-economic activities are very much affected and there is a huge loss over the world in many aspects. If safety measures are not followed strictly in the public places, then there is a rapid spread of the disese at a very faster rate. Hence, this paper provides a thorough survey of the existing computer vision and machine learning-based technological solutions for controlling the spread of the disease. It also discusses some challenges and future perspectives in developing systems for monitoring the COVID-19 safety violations.

19.
8th International Conference on Modelling and Simulation for Autonomous Systems (MESAS) ; 13207:397-416, 2021.
Article in English | Web of Science | ID: covidwho-1819411

ABSTRACT

Most of present humanitarian crises are protracted in nature and their average duration has increased. Climate change, environmental degradation, armed conflicts, terrorism, and migration are producing exponentially growing needs to whom humanitarian organizations are struggling to respond. Novel infectious diseases such as COVID-19 add complexity to protracted crises. Planning to respond to current and future medical threats should integrate terrorist risk assessment, to safeguard population and reduce risks to aid workers. Technologies such as Artificial Intelligence (AI), and Modelling and Simulation (M&S) can play a crucial role. The present research has included the conduct of the United Nations HNPW 2021 session on AI and Medical Intelligence and an exercise on a real scenario. Focusing on medical and terror threats in North East Nigeria operating environment, authors have successfully deployed and tested the Expert.ai Medical Intelligence Platform (MIP) jointly with the MASA SYNERGY constructive simulation, with the aim to improve situational awareness to support decision-making in the context of a humanitarian operation.

20.
Kuwait Journal of Science ; : 30, 2021.
Article in English | Web of Science | ID: covidwho-1819168

ABSTRACT

Coronavirus (COVID-19) has continued to be a global threat to public health. When the coronavirus pandemic began early in 2020, experts wondered if there would be waves of cases, a pattern seen in other virus pandemics. The overall pattern so far has been one of increasing cases of COVID-19 followed by a decline, and we observed a second wave of increased cases and yet we are still exploring this pandemic. Hence, updating the prediction model for the new cases of COVID-19 for different waves is essential to monitor the spreading of the virus and control the disease. Time series models have extensively been considered as the convenient methods to predict the prevalence or spreading rate of the disease. This study, therefore, aimed to apply the Autoregressive Integrated Moving Average (ARIMA) modelling approach for predicting new cases of coronavirus (COVID-19). We propose a deterministic method to predict the basic reproduction number Ro of first and second wave transition of COVID-19 cases in Kuwait and also to forecast the daily new cases and deaths of the pandemic in the country. Forecasting has been done using ARIMA model, Exponential smoothing model, Holt's method, Prophet forecasting model and machine learning models like log-linear, polynomial and support vector regressions. The results presented aligned with other methods used to predict Ro in first and second waves and the forecasting clearly shows the trend of the pandemic in Kuwait. The deterministic prediction of Ro is a good forecasting tool available during the exponential phase of the contagion, which shows an increasing trend during the beginning of the first and second waves of the pandemic in Kuwait. The results show that support vector regression has achieved the best performance for prediction while a simple exponential model without trend gives good optimal results for forecasting of Kuwait COVID-19 data.

SELECTION OF CITATIONS
SEARCH DETAIL