University of Waterloo can actually detect COVID-19 from chest radiographs
New Coronary Pneumonia is still raging, and people all over the world are fighting the epidemic together. A key step in the fight against COVID-19 is effective screening of infected patients so that those infected can receive immediate treatment and care, and are quarantined to reduce the spread of the virus. The standard test is nucleic acid PCR, but the time-consuming operation is complicated. Therefore, the use of artificial intelligence image detection is an option. Recently, the University of Waterloo has released COVID-NET specifically for detecting new coronary pneumonia in lung tablets, with very promising results.
The COVID-19 pandemic continues to have a devastating impact on the health and well-being of the global population. A key step in the fight against COVID-19 is the effective screening of infected patients. One of the most critical screening methods is the use of chest radiography for radiography. Based on this, many artificial intelligence (AI) systems based on deep learning have been proposed, and the results show that it is promising in the accuracy of using chest radiographs to detect COVID-19 infected patients. However, these developed artificial intelligence systems are closed, and the research community cannot further understand and expand them, nor can they access and use the public. Therefore, COVID-Net is introduced, which is a deep convolutional neural network designed for the detection of COVID-19 in chest radiograph images. It is open source and open to the public. The chest radiograph dataset used to train COVID-Net is also described, called COVIDx, which consists of 5941 front and back chest radiograph images from 2839 patients from two open access databases. In addition, it also studies how COVID-net uses interpretable methods to make predictions to gain a deeper understanding of key factors related to COVID cases, thereby helping clinicians improve screening. Never produce a ready solution, hope to open access to COVID-Net, build open source COVIDx datasets as described, leverage, and build highly accurate practical deep learning solutions that are both accelerated by researchers and citizen data scientists COVID-19 cases and expedited treatment of those most in need.
Our company’s X-ray equipment and peripheral equipment such as chest bucky stand can allow doctors to quickly take clear pictures of the lungs, diagnose and treat as soon as possible, and restore health as soon as possible.