Tec engineers attemp to detect COVID-19 with artificial intelligence

Teachers from the School of Engineering and Sciences (EIC) at Tec de Monterrey have developed a system using artificial intelligence that, according to computer calculations, could provide highly effective COVID-19 diagnoses.

This system is based on scanning chest X-rays, and a research protocol involving radiology specialists from TecSalud will be started to validate its results.

The technology works by using deep learning algorithms to develop computational neural networks for identifying characteristic patterns from more than 5,000 X-ray images of negative and positive SARS-CoV-2 cases.

"You often see these algorithms working highly efficiently when applied to facial or voice recognition systems."

"We decided to harness this technology to identify the main descriptors from X-ray images taken of people with COVID-19," explained Sergio Uribe, Director of the Center for Innovation in Design and Technology at the Tec.

This system also uses machine learning to identify the information processed and provide a diagnosis to confirm or rule out the incidence of coronavirus in each case, added this professor.

The prototype already operates with a training level of sensitivity and specificity that is higher than 95 percent. These are two important parameters in methods of clinical diagnosis for distinguishing patients suffering from the disease from those who aren't.

Sergio Uribe, who leads this project alongside Professor Cristian Mendoza, said that although this method has certain advantages over PCR tests, such as immediate diagnosis and a considerably lower cost, it should be considered a complementary method to the standard test.

YOU'LL SURELY WANT TO READ THIS TOO: