COVID-Net, trained with 5,941 images of 2,839 patients with various lung diseases, has been publicly launched for the rest of the world’s researchers to perfect it until it is able to reliably diagnose the disease, as has already been done with other ailments.
What is it? COVID-Net is a convolutional neural network, a type of AI that is particularly good at recognizing images. Developed by Linda Wang and Alexander Wong at the University of Waterloo and IA firm DarwinAI in Canada, COVID-Net was trained to identify signs of Covid-19 on chest radiographs using 5,941 images taken from 2,839 patients with various lung conditions, including bacterial infections, non-Covid viral infections as well as Covid-19 infections. The dataset is provided along with the tool so that researchers, or anyone who wants to play, can explore and modify it.
Don’t believe the hype: Several research teams have announced AI tools that can diagnose Covid-19 from X-rays in recent weeks. But none have been made available to the public, making it difficult to assess their accuracy. DarwinAI is taking a different approach. He notes that COVID-Net is “by no means a production-ready solution” and encourages others to help become one. DarwinAI, whose CEO Sheldon Fernández will speak at EmTech Digital tomorrow, also wants the tool to explain his reasoning, making it easier for healthcare workers to use.
COVID-Net has yet to be tested, but follows in the footsteps of a previous success story. Many of the breakthroughs in computer vision in the past 10 years are due to the public launch of ImageNet, a large data set of millions of everyday images, and AlexNet, a trained convolutional neural network. Investigators have been building on both since then.