Performance Evaluation of the NASNet Convolutional Network in the Automatic Identification of COVID-19

Abstract

This paper evaluates the performance of the Neural Architecture Search Network (NASNet) in the automatic detection of COVID-19 (Coronavirus Disease 2019) from chest x-ray images. COVID-19 is a disease caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) that produces in patients fever, cough, shortness of breath, muscle pain, sputum production, diarrhea, and even sore throat. The virus spreads through the air, and to date, is expanding as a global pandemic. There is no vaccine, and it is fatal to approximately 2-7% of the infected population. Among the clinical and paraclinical characteristics of infected patients, nodules have been identified in images of chest x-rays that can be visually identified, producing a simple, rapid, and generally available method of identification. However, the rapid spread of the disease means that there is a lack of specialized medical personnel capable of identifying it, which is why automated schemes are being developed. We propose the tuning of a NASNet-type convolutional model to automatically determine the initial state of a patient in the triage process or intervention protocol of health care centers. The neural network is trained with public images of cases positively identified as patients infected with the virus and patients in normal conditions without infection. Performance evaluation is also done with real images unknown to the neuronal model. As for performance metrics, we use the function of loss of cross-entropy (categorical cross-entropy), the accuracy (or success rate), and the MSE (Mean Squared Error). The tuned model was able to correctly classify the test images with an accuracy of 97%.

Light-weighted ensemble network with multilevel activation visualization for robust diagnosis of COVID19 pneumonia from large-scale chest radiographic database

“…In this section, a comprehensive comparison of the proposed method is made with existing state-of-the-art deep feature-based CAD methods related to COVID19 diagnostics [13] , [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] , [25] . Most of these methods used the existing pretrained networks and applied the end-to-end transfer learning approach for the automated diagnosis of COVID19 infection.…”

“…However, these comparative studies used limited radiographic datasets and different experimental protocols. For a fair comparison, the quantitative results of these baseline methods [13] , [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] , [25] were assessed based on the selected datasets and experimental protocol. In details, the pretrained backbones of the baseline methods [13] , [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] , [25] were selected and fine-tuned with the selected datasets.…”

“…In details, the pretrained backbones of the baseline methods [13] , [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] , [25] were selected and fine-tuned with the selected datasets. The training parameters were chosen with the experimental data during the fine-tuning of backbones of the baseline methods [13] , [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] , [25] , and same method was adopted for the proposed model. Additionally, the same experimental protocol (including five-fold cross-validation) was adopted for the performance comparisons of the baseline methods [13] , [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] , [25] and the proposed model.…”

“… [17] VGG19 139.58 91.54 91.33 92.26 90.47 94.54 95.30 95.39 95.14 95.65 97.84 Martínez et al. [25] NASNet 4.27 93.68 93.49 94.19 92.82 96.67 94.06 94.06 93.89 94.23 97.07 Misra et al. [24] ResNet18 11.18 92.96 92.76 93.41 92.14 95.06 95.59 95.69 95.44 95.95 97.79 Farooq et al.…”

“…Recent advancements in artificial intelligence (AI) have shown a remarkable breakthrough in the development of computer-aided diagnosis (CAD) tools [6] , [7] , [8] , [9] , [10] , [11] , [12] , [13] , [14] , [15] , [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] , [25] , [26] , [27] . In particular, with the advent of deep learning algorithms, such automated tools have shown significant decision-making performance in different medical domains, including radiology.…”