Predicting COVID-19 from Chest X-ray Images using a New Deep Learning Architecture

The spread of coronavirus disease in late 2019 caused huge damage to human lives and forced a chaos in health care systems around the globe. Early diagnosis of this disease can help separate patients from healthy people. Therefore, precise COVID-19 detection is necessary to prevent the spread of this virus. Many artificial intelligent technologies for example deep learning models have been applied successfully for this task by employing chest X-ray images. In this paper, we propose to classify chest X-ray images using a new end-To-end convolutional neural network model. This new model consists of six convolutional blocks. Each block consists of one convolutional layer, one ReLU layer, and one max-pooling layer. The new model was applied on a challenging imbalanced COVID19 dataset of 5000 images, divided into two classes, COVID and Non-COVID. In experiments, the input image is first resized to 256×256×3 before being fed to the model. Two metrics were used to test our new model: sensitivity and specificity. A sensitivity rate of 97% was achieved along with a specificity rate of 99.32%. These results are promising when compared to other deep learning models applied on the same dataset. © 2022 IEEE.

Prediction of covid-19 from chest X-ray images using multiresolution texture classification with robust local features

The COVID-19 contagious disease that spread around the world, have a huge risk on people and already caused millions of deaths forcing a global pandemic in 2020. Diagnosing patients with this disease is very critical allowing fast care response and to isolate them from public. As the virus spread widely to millions of people, the fastest way to detect it is by analyzing radiology images. Early studies showed irregularity in the chest X-ray images of patients with high clinical belief of COVID-19 infection. Hence, these studies motivated us to investigate the use of machine learning techniques to help diagnosing COVID-19 patients from chest CT scans. In this paper, we propose to use a robust feature extraction descriptor and to apply a Random Forests classifier to predict COVID-19 disease in a dataset of 5000 images. First, 408 texture features are extracted using a powerful variation of Local Binary Patterns descriptor called Rotation Invariant Co-occurrence among Local Binary Patterns. Then, Random Forests classifier is applied with 250 trees to perform the classification task. Moreover, the performance of our approach was improved by using a multiresolution scheme where features are extracted from both the original input image and the subsampled image. Two metrics were used to evaluate our approach, sensitivity and specificity. We achieved 99.0% and 91.3% for both metrics, respectively. Our results are close to the state-of-the-art deep learning methods on the same dataset. © 2021 IEEE.