The COVID-19 spread across the world has created a sense of panic among all. The reason for this virus to become pandemic is it being asymptomatic for most of the people. In this whole confusion of who has an infection and who doesn't. Developers are trying their bit by creating new applications and software to help the government.
Artificial Intelligence (AI) can help fight COVID-19, using applications, including population screening, sending a notification when to seek medical help, and tracking how the infection has spread. One such tool is COVID-Net AI that can help, in identifying COVID-19 through chest X-Ray, developed by Canadian Startup Darwin AI.
They are using Convolution Neural Network(CNN) which is a deep learning algorithm for image recognition, a CNN is able to capture the spatial and temporal dependency in an image using optimum filters. The input image is converted into a matrix of RGB Image and CNN is applied by which image recognition can be done, and it can classify the patient with COVID-19 or not as shown in Fig.2. The algorithm was trained with 5,941 chest x-rays of 2,839 patients, all having different symptoms, some shown have COVID-19 symptoms while some don't have COVID-19 symptoms, and were successful in the diagnosis of COVID-19 cases. Researchers have made this tool public so that researchers from other countries and doctors can help fine-tune it further and help detect COVID-19 cases faster. While many companies have announced AI tools to diagnose COVID-19, none have made it public as of yet.
As cases are rising rapidly across the world, this tool might be revolutionary in combating with COVID-19 pandemic.
ConversionConversion EmoticonEmoticon