COVID-CheXNet: hybrid deep learning framework for identifying COVID-19 virus in chest X-rays images

Abstract

The outbreaks of Coronavirus (COVID-19) epidemic have increased the pressure on healthcare and medical systems worldwide. The timely diagnosis of infected patients is a critical step to limit the spread of the COVID-19 epidemic. The chest radiography imaging has shown to be an effective screening technique in diagnosing the COVID-19 epidemic. To reduce the pressure on radiologists and control of the epidemic, fast and accurate a hybrid deep learning framework for diagnosing COVID-19 virus in chest X-ray images is developed and termed as the COVID-CheXNet system. First, the contrast of the X-ray image was enhanced and the noise level was reduced using the contrast-limited adaptive histogram equalization and Butterworth bandpass filter, respectively. This was followed by fusing the results obtained from two different pre-trained deep learning models based on the incorporation of a ResNet34 and high-resolution network model trained using a large-scale dataset. Herein, the parallel architecture was considered, which provides radiologists with a high degree of confidence to discriminate between the healthy and COVID-19 infected people. The proposed COVID-CheXNet system has managed to correctly and accurately diagnose the COVID-19 patients with a detection accuracy rate of 99.99%, sensitivity of 99.98%, specificity of 100%, precision of 100%, F1-score of 99.99%, MSE of 0.011%, and RMSE of 0.012% using the weighted sum rule at the score-level. The efficiency and usefulness of the proposed COVID-CheXNet system are established along with the possibility of using it in real clinical centers for fast diagnosis and treatment supplement, with less than 2 s per image to get the prediction result. Keywords Coronavirus COVID-19 epidemic Á Deep learning Á Transfer learning Á ResNet34 model Á Chest radiography imaging Á Chest X-ray images Communicated by Valentina E. Balas.

Classification of COVID-19 electrocardiograms by using hexaxial feature mapping and deep learning

“…Recently, presented studies to automatically diagnose COVID-19 with deep learning have emphasized well-known architecture ResNet [ 72 ]. Accordingly, in addition to the experiments performed, generated hexaxial mapping images were trained with ResNet-50 architecture to compare with our proposed architecture.…”

“…Additionally, many deep learning-based studies have used radiographic images for the detection of COVID-19 and many of them have achieved outstanding classification performance. The following studies can be shown as an example: Al-Waisy et al [ 72 ] achieved accuracy of 99.99%, Dhiman et al [ 79 ] achieved accuracy of 98.54%, Ozturk et al [ 14 ] achieved accuracy of 98.08%, and Ahuja et al [ 74 ] achieved accuracy of 99.4%. The main reason for the success of the mentioned studies is that the most common symptom of COVID-19 disease is lung involvement [ 80 ] and the symptoms can be clearly observed on radiographic lung images [ 81 ].…”

“…For this purpose, hexaxial feature mapping images were trained with the network architectures which are suggested by recent state-of-the-art studies. When the classification results are compared ResNet-50 [ 72 ], AlexNet [ 73 ], ResNet-8 [ 16 ], and SqueezeNet [ 74 ] yielded an accuracy of 73.65%, 93.60%, 85.12%, and 92.52%, respectively. The results revealed that AlexNet which has less network depth achieved significant accuracy compared to well-known architectures.…”

Chest Imaging of Patients with Sarcoidosis and SARS-CoV-2 Infection. Current Evidence and Clinical Perspectives

“…Therefore, novel algorithms could be designed to evaluate the probability of predominant impairment from sarcoidosis or COVID-19, according to predefined scores [ 60 ]. Such models could represent the key for distinction in the diagnostic approach to the conditions and a correct identification could be important to orient towards a correct therapy.…”

Efficient Detection of Knee Anterior Cruciate Ligament from Magnetic Resonance Imaging Using Deep Learning Approach

“…Nowadays, researchers are using deep learning with a model of CNN and its architectures in several applications. The CNNs architectures have an input layer and an output layer, and there are also many convolutional layers, pooling layers, rectified linear unit layers, dense layers and dropout layers [ 17 , 18 ]. The CNN shows huge success in the analysis of radiography X-rays in the knee osteoarthritis automatically, as there is no need of image pre-processing [ 19 , 20 ].…”