Deep Transfer Learning Based Unified Framework for COVID19 Classification and Infection Detection from Chest X-Ray Images

2021 5th International Conference on Computing Methodologies and Communication (ICCMC), 2021

Indonesian Journal of Electrical Engineering and Computer Science, 2022

The novel coronavirus, also known as COVID-19, initially appeared in Wuhan, China, in December 2019 and has since spread around the world. The purpose of this paper is to use deep convolutional neural networks (DCCN) to improve the detection of COVID-19 from X-ray images. In this study, we create a DCNN based on a residual network (Resnet-50) that can identify COVID-19 from two other classes (pneumonia and normal) in chest X-ray images. DCNN was evaluated using two classification methods: binary (BC-1: COVID-19 vs. normal, BC-2: COVID-19 vs. pneumonia) and multi-class (pneumonia vs. normal vs. COVID-19). In all experiments, four fold cross-validation was used to train and test the model. This architecture's average accuracy is 99.9% for BC-1, 99.8% for BC-2, and 97.3% for multiclass cases. The experimental findings demonstrated that the suggested system detects COVID-19 with an average precision and sensitivity of 95% and 95.1% for multi-class classification, respectively. According to our findings, the proposed DCNN may help health professionals in confirming their first evaluation of COVID-19 patients.

Elsevier, 2022

COVID-19 cases are putting pressure on healthcare systems all around the world. Due to the lack of available testing kits, it is impractical for screening every patient with a respiratory ailment using traditional methods (RT-PCR). In addition, the tests have a high turnaround time and low sensitivity. Detecting suspected COVID-19 infections from the chest X-ray might help isolate high-risk people before the RT-PCR test. Most healthcare systems already have X-ray equipment, and because most current X-ray systems have already been computerized, there is no need to transfer the samples. The use of a chest X-ray to prioritize the selection of patients for subsequent RT-PCR testing is the motivation of this work. Transfer learning (TL) with fine-tuning on deep convolutional neural network-based ResNet50 model has been proposed in this work to classify COVID-19 patients from the COVID-19 Radiography Database. Ten distinct pre-trained weights, trained on varieties of large-scale datasets using various approaches such as supervised learning, self-supervised learning, and others, have been utilized in this work. Our proposed 2021_ _ _1 model, pre-trained on the iNat2021 Mini dataset using the SwAV algorithm, outperforms the other ResNet50 TL models. For COVID instances in the two-class (Covid and Normal) classification, our work achieved 99.17% validation accuracy, 99.95% train accuracy, 99.31% precision, 99.03% sensitivity, and 99.17% F1-score. Some domain-adapted (_ ℎ − 14) and in-domain (ChexPert, ChestX-ray14) models looked promising in medical image classification by scoring significantly higher than other models.

International Journal For Research In Applied Science & Engineering Technology, 2020

The ongoing novel corona virus has spread all over the world and became a pandemic. This pandemic situation has led to a major crisis in healthcare systems and the global economy. As Covid-19 positive patient's increasing day by day, the crucial task is to detect and monitor disease efficiently and facilitate the results of Covid-19 positive patients to cure them as soon as possible. Currently used RT-PCR (Reverse transcription-polymerase chain reaction) testing method act as a goldmine for detecting Covid-19. But the total turnaround time required for Disease diagnosis is very large. This long turnaround time sometimes leads to patient deaths. To avoid that and detecting Covid-19 positive patients in a less time ,author proposed a method in this paper that uses Chest x-ray images for patient diagnosis and disease classification. Deep learning architecture called Convolutional neural network helps in diagnosis of patient. The tremendous success of the Convolutional neural network at image processing tasks in recent years extremely increased the use of electronic medical records and diagnostic imaging. To train and test the neural model the paper used a publicly available dataset that contains COVID-19, pneumonia, and normal patient Chest X-ray images. Also for experimental analysis a CovidNet20, Convolutional architecture was developed for disease classification along with transfer learning DenseNet121 pretrained model used for training and testing of the classification model. The proposed model able to differentiate COVID-19 and normal images as binary classification with 100% and 99% accuracy on DenseNet121 and CovidNet20 model. And, on multiclass classification with COVID-19, Normal and Pneumonia as classes Densenet121 gives 97% and CovidNet20 gives 98% accuracy.

International Journal of Computer Applications, 2020

Computers, Materials & Continua

2021 5th International Conference on Intelligent Computing and Control Systems (ICICCS), 2021

Today SARS-COVID-2 causes Novel Coronavirus diseases throughout in more than 150 countries all over the world. The quicker diagnosis is very crucial to reduce the outbreak of this diseases. The clinic al studies regarding this disease has shown that patients lungs are very much affected after the infection of coronavirus. Chest X-Ray, CT Scan are the most effective imaging approaches for identification of COVID 19 disease. Deep Learning approaches are one of the important approaches of machine learning that gives a critical analysis regarding for study of large amount of image datasets that can make some earlier impact of diseases. in recent years. To analyze the disease 1000 images are used for training and 150 images are used for testing the data from an online available standardized dataset of Kaggle. Here the images are taken as Covid and Non-Covid as the 2 class levels to classify the images using CNN. Here the activation function ReLU provides more than 90 percent of accuracy ...

2021

With the ongoing outbreak of the COVID-19 global pandemic, the research community still struggles to develop early and reliable prediction and detection mechanisms for this infectious disease. The commonly used RT-PCR test is not readily available in areas with limited testing facilities, and it lags in performance and timeliness. This paper proposes a deep transfer learning-based approach to predict and detect COVID-19 from digital chest radiographs. In this study, three pre-trained convolutional neural network-based models (VGG16, ResNet18, and DenseNet121) have been fine tuned to detect COVID-19 infected patients from chest X-rays (CXRs). The most efficient model is further used to identify the affected regions using an unsupervised gradient-based localization technique. The proposed system uses a classification approach (normal vs. COVID-19 vs. pneumonia vs. lung opacity) using three supervised classification algorithms followed by gradient-based localization. The training, vali...

SN Computer Science

Multimedia Tools and Applications