MobApp4InfectiousDisease: Classify COVID-19, Pneumonia, and Tuberculosis

Illness due to infectious diseases has been always a global threat. Millions of people die per year due to COVID-19, pneumonia, and Tuberculosis (TB) as all of them infect the lungs. For all cases, early screening/diagnosis can help provide opportunities for better care. To handle this, we develop an application, which we call MobApp4InfectiousDisease that can identify abnormalities due to COVID-19, pneumonia, and TB using Chest X-ray image. In our MobApp4InfectiousDisease, we implemented a customized deep network with a single transfer learning technique. For validation, we offered in-depth experimental study and we achieved, for COVID-19-pneumonia-TB cases, accuracy of 97.72%196.62%199.75%, precision of 92.72%1100.0%199.29%, recall of 98.89%188.54%199.65%, and F1-score of 95.00%194.00%199.00%. Our results are compared with state-of-the-art techniques. To the best of our knowl-edge, this is the first time we deployed our proof-of-the-concept MobApp4InfectiousDisease for a multi-class infec-tious disease classification. © 2022 IEEE.

COVID-19 Detection Using Chest X-Ray Images with a RegNet Structured Deep Learning Model

AI-based medical image processing has made significant progress, and it has a significant impact on biomedical research. Among the imaging variants, Chest x-rays imaging is cheap, simple, and can be used to detect influenza, tuberculosis, and various other illnesses. Researchers discovered that coronavirus spreads through the lungs, causing severe injuries during the COVID19 pandemic. As a result, chest x-rays can be used to detect COVID-19, making it a more robust detection method. In this paper, a RegNet hierarchical deep learning-based model has been proposed to detect COVID-19 positive and negative cases using CXI. The RegNet structure is designed to develop a model with a small number of epochs and parameters. The performance measurement found that the model takes five periods to reach a total accuracy of 98.08%. To test the model, we used two sets of data. The first dataset consists of 1200 COVID-19 positive CXRs and 1,341 COVID-19 negative CXRs, and the second dataset consists of 195 COVID-19 positive CXRs and 2,000 COVID-19 negative CXRs;all of these are publicly available. We obtained precision of 99.02% and 97.13% for these datasets, respectively. As a result of this finding, the proposed approach could be used for mass screening, and, as far as we are aware, the results achieved indicate that this model could be used as a screen guide. © 2021, Springer Nature Switzerland AG.