Using Deep Learning in X-rays to detect coronavirus

Perhaps, new technologies are one of the most useful resources in the fight against COVID-19. Artificial Intelligence is among the most involved technologies in the battle against the virus.

It has proven to be particularly useful for monitoring and managing the virus. Its many applications have turned it into a containment mechanism intended to put a stop to the pandemic. However, while we fight against the virus we must continue to search for new applications of AI.

One of the applications that will help in the fight against Covid-19 is image processing. Deep Learning has been implemented in a system that detects patients with Covid-19 through chest X-rays. Thus, this model, created by the Institute of Knowledge Engineering (ICC), is capable of discerning between healthy and sick patients with an accuracy of over 90%.

This type of model learns on the basis of similar data. For this system's learning, three sources were used with radiographs of patients that had to be classified in three phases: training, validation of progress and test of effectiveness involving new images.

Thanks to this model, it will be possible to detect pneumonia and problems derived from coronavirus much more quickly, thus complementing "traditional" virus diagnoses.

At Air Institute we strive towards the development of new technologies such as Artificial Intelligence and we search for new solutions to problems in numerous sectors, such as the health sector.