Comparative Analysis of Deep Learning Models for Covid-19 Detection from Chest X-rays

Lungs are the organs which play key role in human respiratory system. The severity of infections caused to the lungs might vary from mild to moderate. Chest X-Ray is a principal diagnostic tool used in detecting various types of lung diseases. The whole world is struggling due to a pandemic arised in 2019, known as Coronavirus disease or Covid-19, a severe respiratory infection. The medical industry demanded the use of computer aided techniques for analysing extremity of the disease. This work aims to examine the effectiveness of pretrained deep learning models in classifying chest X-rays as Covid, Viral pneumonia and Healthy cases. We have used largest publicly accessible Covid dataset, QaTa Cov-19 for conducting experiments. Out of six fine tuned deep learning pretrained network models, Densenet 201 outperformed with highest accuracy of 98.6% and AUC of 0.9996. © 2022 IEEE.

Systematic Analysis of Several Deep Learning Approaches for COVID-19 Detection Using X-ray Images

COVID-19 is a virus-borne malady. A clinical study of infected COVID-19 patients found that most COVID-19 patients suffered lung infection after contracting the disease. Consequently, chest X-rays are a more effective and lower-cost imaging technique for diagnosing lung-related problems. This study used deep learning models, including MobileNetV2,DenseNet201, ResNet50, and VGG19, for COVID-19 prediction. For the study, we used chest X-ray image data for binary classification of COVID-19. 7207 chest X-ray image data were obtained from the Kaggle repository, with 5761 being utilized for training and 1446 being used for validation. A comparative analysis was conducted among the models and examined their accuracy. It has been determined that the DenseNet201 models achieved the highest accuracy of 93.02% for detecting COVID-19 in the lowest compilation time of 27secs. The models, MobileNetV2, ResNet50, and VGG19 had the accuracy rate of 77.28%, 65.86% and 74.92%, respectively. The research indicates that the DenseNet201 model is the most effective in detecting COVID-19 using x-ray imaging. © 2022 IEEE.

COVID-19 Detection Through Deep Learning Algorithms using Chest X-ray Images

COVID-19 became a pandemic affecting the lives of every human globally by the end of 2019. The disease impaired the lungs of infected patients. Precise prediction and diagnosis of COVID-19 disease are challenging due to its resemblance to viral pneumonia. Using multiple deep learning approaches, the researchers used chest X-ray (CXR) imaging to diagnose COVID-19. The X-ray image dataset from Kaggle is used for the study by selecting the COVID-19 and normal class. InceptionV3, MobileNetV2, VGG19,VGG16 and ResNet50 are the five neural networks used for binary classification of COVID-19. The accuracy of MobileNetV2 surpasses that of the remainder of the model by 93.02%. However, it has a compilation time of 1836 seconds per epoch. Besides, VGG16 has an accuracy of 92.37%, with a compilation time of 603 seconds per epoch. Compared to these models, Inceptonv3, Resnet50 and VGG19 perform with an accuracy score of 86.42%, 68.34% and 91.79%. Applying deep learning techniques to COVID-19 radiological imaging holds great promise for enhancing the accuracy of diagnosis when in comparison to the gold standard RT-PCR test and assisting healthcare professionals in making decisions quickly © 2022 IEEE.

Deep learning-based hybrid models for prediction of COVID-19 using chest X-ray

The ongoing COVID-19 virus infection has ended up being the biggest pandemic to hit mankind in the last century. It has infected in excess of 50 Million across the globe and has taken in excess of 1.5 Million lives. It has posed problems even to the best healthcare systems across the globe. The best way to reduce the spread and damage of COVID-19 is by early detection of the infection and quarantining the infected patients with necessary medical care. COVID-19 infection can be detected by a chest X-ray. With limited rapid COVID-19 testing kits, this approach of detection with the aid of deep learning can be adopted. The only problem being, the side effects of COVID-19 infection imitate those of conventional Pneumonia, which adds some complexity in utilizing the Chest X-rays for its prediction. In this investigation, we attempt to investigate four approaches i.e., Feature Ensemble, Feature Extraction, Layer Modification and weighted Max voting utilizing State of the Art pre-trained models to accurately identify between COVID-19 Pneumonia, Non-COVID-19 Pneumonia, and Healthy Chest X-ray images. Since very few images of patients with COVID-19 are publicly available, we utilized combinations of image processing and data augmentation methods to build more samples to improve the quality of predictions. Our best model i.e., Modified VGG-16, has achieved an accuracy of 99.5216%. More importantly, this model did not predict a False Negative Normal (i.e., infected case predicted as normal), making it the most attractive feature of the study. The establishment of such an approach will be useful to predict the outbreak early, which in turn can aid in controlling it effectively. © 2022 Elsevier Inc. All rights reserved.

Efficient Grad-Cam-Based Model for COVID-19 Classification and Detection

Corona Virus (COVID-19) is a novel virus that crossed an animal-human barrier and emerged in Wuhan, China. Until now it has affected more than 119 million people. Detection of COVID-19 is a critical task and due to a large number of patients, a shortage of doctors has occurred for its detection. In this paper, a model has been suggested that not only detects the COVID-19 using X-ray and CT-Scan images but also shows the affected areas. Three classes have been defined;COVID-19, normal, and Pneumonia for X-ray images. For CT-Scan images, 2 classes have been defined COVID-19 and non-COVID-19. For classification purposes, pre-trained models like ResNet50, VGG-16, and VGG19 have been used with some tuning. For detecting the affected areas Gradient-weighted Class Activation Mapping (GradCam) has been used. As the X-rays and ct images are taken at different intensities, so the contrast limited adaptive histogram equalization (CLAHE) has been applied to see the effect on the training of the models. As a result of these experiments, we achieved a maximum validation accuracy of 88.10% with a training accuracy of 88.48% for CT-Scan images using the ResNet50 model. While for X-ray images we achieved a maximum validation accuracy of 97.31% with a training accuracy of 95.64% using the VGG16 model. © 2023 CRL Publishing. All rights reserved.

Small-Scale CNN-N model for Covid-19 Anomaly Detection and Localization From Chest X-Rays

Covid-19 has been posing a serious challenge to scientists and health organizations around the world in terms of detection and its treatment. Common methods are CT-Scans and X-rays to analyze the images of lungs for COVID-19. These days diagnosing covid-19 by manually looking at the reports has become difficult and challenging in the pandemic. Pneumonia and pulmonary infections along with covid-19 cause inflammation and fluids in the lungs. Covid-19 X-rays are very similar to viral and bacterial Pneumonia X-rays. So it becomes very difficult to differentiate between covid-19 and Pneumonia. In this paper we propose a computer vision model to detect the presence of covid19 infection along with the localization of the infection in the lungs. 6337 images consisting of Negative for pneumonia, Typical Appearance, Intermediate Appearance and Atypical Appearance is considered. Although there are pre-trained CNN models which perform well on the data, this paper aims at reducing the size of the model and validate its performance on other datasets. Different image sizes are also considered. A small scale CNN model is built from scratch to detect and localize covid-19 abnormalities on chest radiographs using object detection algorithms like Yolov5 with different weights. There is a significant reduction in model size and parameters compared to many state of the art pre-trained models thereby ensuring efficient detection of covid-19 anomalies and show the region of infection to ensure timely treatment before it causes severe infection. © 2022 IEEE.

COKG-QA: Multi-hop Question Answering over COVID-19 Knowledge Graphs

COVID-19 evolves rapidly and an enormous number of people worldwide desire instant access to COVID-19 information such as the overview, clinic knowledge, vaccine, prevention measures, and COVID-19 mutation. Question answering (QA) has become the mainstream interaction way for users to consume the ever-growing information by posing natural language questions. Therefore, it is urgent and necessary to develop a QA system to offer consulting services all the time to relieve the stress of health services. In particular, people increasingly pay more attention to complex multi-hop questions rather than simple ones during the lasting pandemic, but the existing COVID-19 QA systems fail to meet their complex information needs. In this paper, we introduce a novel multi-hop QA system called COKG-QA, which reasons over multiple relations over large-scale COVID-19 Knowledge Graphs to return answers given a question. In the field of question answering over knowledge graph, current methods usually represent entities and schemas based on some knowledge embedding models and represent questions using pre-trained models. While it is convenient to represent different knowledge (i.e., entities and questions) based on specified embeddings, an issue raises that these separate representations come from heterogeneous vector spaces. We align question embeddings with knowledge embeddings in a common semantic space by a simple but effective embedding projection mechanism. Furthermore, we propose combining entity embeddings with their corresponding schema embeddings which served as important prior knowledge, to help search for the correct answer entity of specified types. In addition, we derive a large multi-hop Chinese COVID-19 dataset (called COKG-DATA for remembering) for COKG-QA based on the linked knowledge graph OpenKG-COVID19 launched by OpenKG((1)), including comprehensive and representative information about COVID-19. COKG-QA achieves quite competitive performance in the 1-hop and 2-hop data while obtaining the best result with significant improvements in the 3-hop. And it is more efficient to be used in the QA system for users. Moreover, the user study shows that the system not only provides accurate and interpretable answers but also is easy to use and comes with smart tips and suggestions.

An AI-enabled pre-trained model-based Covid detection model using chest X-ray images.

The year 2020 and 2021 was the witness of Covid 19 and it was the leading cause of death throughout the world during this time period. It has an impact on a large geographic area, particularly in countries with a large population. Due to the fact that this novel coronavirus has been detected in all countries around the world, the World Health Organization (WHO) has declared Covid-19 to be a pandemic. This novel coronavirus spread quickly from person to person through the saliva droplets and direct or indirect contact with an infected person. The tests carried out to detect the Covid-19 are time-consuming and the primary cause of rapid growth in Covid19 cases. Early detection of Covid patient can play a significant role in controlling the Covid chain by isolation the patient and proper treatment at the right time. Recent research on Covid-19 claim that Chest CT and X-ray images can be used as the preliminary screening for Covid-19 detection. This paper suggested an Artificial Intelligence (AI) based approach for detecting Covid-19 by using X-ray and CT scan images. Due to the availability of the small Covid dataset, we are using a pre-trained model. In this paper, four pre-trained models named VGGNet-19, ResNet50, InceptionResNetV2 and MobileNet are trained to classify the X-ray images into the Covid and Normal classes. A model is tuned in such a way that a smaller percentage of Covid cases will be classified as Normal cases by employing normalization and regularization techniques. The updated binary cross entropy loss (BCEL) function imposes a large penalty for classifying any Covid class to Normal class. The experimental results reveal that the proposed InceptionResNetV2 model outperforms the other pre-trained model with training, validation and test accuracy of 99.2%, 98% and 97% respectively.

Deep Learning for Real-Time Face Mask Detection

The corona virus (COVID19) pandemic requires immediate action to avoid adverse effects on local health and the global economy. Due to the effects of COVID19, most of the people lives have been reversed. In the absence of effective antivirals and inadequate medical resources, UN agencies propose a number of measures to regulate infection rates and prevent the limited medical resources from being exhausted. Wearing a face mask is a type of the non-pharmaceutical intervention techniques that can block the primary care of viral droplets ejected by an infected person. According to government basics, it is important for everyone in every country to wear a mask. The government recommends wearing a mask, but many do not. Mask detection is very important in this situation. To contribute to community health, this study aims to develop highly accurate and timely techniques for detecting non-face masks in public and encouraging people to use them. It is said that "Increase the number of people who wear masks correctly and reduce the number of infected people". Starting with MobileNet V2 as a baseline, we used the concept of transfer learning to fuse high levels of linguistic data during mask recognition. For the face detection module, we used Caffe Model in conjunction with OpenCV's DNN module. The anticipated model's remarkable performance makes it ideal for live video police work equipment that detects face masks in real - time. © 2022 IEEE.

Transfer Learning Techniques with SVM For Covid-19 Disease Prediction Based On Chest X-Ray Images

Image detection and classification based on X-ray imaginary has been a trending field of study related to medical diagnosis. The challenges associated with diagnosing infections using X-ray images have become a drawback in the medical field. Among the detection of infections, COVID-19 can be considered a novel disease that has undergone immense movement in the medical field during the past couple of years while becoming a severe problem to the world community. Moreover, the diagnosis of COVID-19 related to the radiological manifestations using chest X-ray images is unfamiliar since it is a new experience for many experts. Therefore, this study is carried out to predict COVID-19 with the use of chest X-ray images done under binary and multiclass classifications. For the proposed work, a model using the pre-trained models VGG16, VGG19, InceptionV3, MobileNetV2, and Xception and whereas the Support Vector Machine (SVM) was implemented as the end layers and the model fitting was done using Adam optimizer along with the learning rate of 0.0004. The study used a dataset selected from the Kaggle repository, considered one of the benchmark datasets. An accuracy of 100% was obtained by the proposed model for training, testing, and validation in the task of binary classification of COVID-19 and normal classes while being capable of outperforming 97% accuracy for training, 99% accuracy for both testing and validation purposes under the classification of multi-classes namely COVID-19, pneumonia and normal classes using the proposed model where VGG16 outperformed as the base model. Overall, this study was contributed to overcoming the vital yet challenging task of detecting COVID-19 from other infections related to lungs using chest X-ray images. © 2022 IEEE.

Deep Learning Based Mathematical Model for Feature Extraction to Detect Corona Virus Disease using Chest X-ray Images

Currently, the entire world is fighting against the Corona Virus (COVID-19). As of now, more than thirty lacs of people all over the world were died due to the COVID-19 till April 2021. A recent study conducted by China suggests that Chest CT and X-ray images can be used as a preliminary test for COVID detection. This paper propose a transfer learning-based mathematical COVID detection model, which integrates a pre-trained model with the Random Forest Tree (RFT) classifier. As the available COVID dataset is noisy and imbalanced so Principal Component Analysis (PCA) and Generative Adversarial Networks (GANs) is used to extract most prominent features and balance the dataset respectively. The Bayesian Cross-Entropy Loss function is used to penalize the false detection differently according to the class sensitivity (i.e., COVID patient should not be classified as Normal or Pneumonia class). Due to the small dataset, a pre-trained model like VGGNet-19, ResNet50 and Inception_ResNet_V2 were chosen to extract features and then trained them over the RFT for the classification task. The experiment results showed that ResNet50 gives the maximum accuracy of 99.51%, 98.21%, and 97.2% for training, validation, and testing phases, respectively, and none of the COVID Chest X-ray images were classified as Normal or Pneumonia classes.