Wednesday, August 4

Nerea Luis: «Artificial intelligence can accelerate medical decision-making in a pandemic»


The Madrid engineer, an expert in “machine learning” and “computer vision” projects, was awarded in 2019 by the Royal House with the Order of Civil Merit and is one of the top women leaders in Spain in the revelation category. It belongs to the “Los 100 de Cotec” network.

What moment is artificial intelligence going through?

Today it is a very transversal field. It started out deeply rooted on the research side. It was difficult for algorithms to work in real fields, but after the development of the cloud and the arrival of companies that work with artificial intelligence modules, it is much more accessible. Since the boom of 2016 until now it has expanded to multiple areas. On a daily basis, we consume it when we enter e-commerce sites or watch series on television platforms. These services already have artificial intelligence modules based on recommendation systems implemented. When we use Google Maps we have trained models regarding time, schedule or type of transportation. There are projects in medicine or cinema, for example. Artificial intelligence can act as a catalyst for all these sectors. Algorithms are very transversal and we have not yet seen their full potential.

How can artificial intelligence help prevent a new pandemic?

It is difficult for it to foresee new pandemics but it is capable of alerting if it is trained with a multitude of data related to viral and immunological issues. All that part is being worked on in a very embryonic way. There are alternatives to “machine learning” such as federated learning that will increasingly encourage this knowledge to expand sooner.

“Robots can play a big role in collaborative tasks such as hospital disinfection”


decoration

What does this federated learning consist of?

In the case of covid, proposals have been made from the point of view of research but taking them to the real field is difficult because it takes a long time and in Medicine you have to certify everything to provide security and reliability to what you do. It does study how to share that knowledge in the future. Faced with a new disease, doctors need diverse approaches such as federated learning. This means having a system that allows, beyond understanding the disease in full, each medical team to train its own models, sharing out the knowledge itself, the output generated by having a history of patients with the same symptoms that They have motivated a series of decisions, from a basic triage to something more elaborate. That model is sent to another hospital in that federated network that is generated. This would make it easier for hospitals in the United Kingdom and Spain, for example, to exchange their discoveries by excluding patient data for confidentiality so that artificial intelligence works in privacy.

How does it apply to covid?

In pilot mode. These lines are very interesting beyond the typical decision-making systems so that they can become global and knowledge can be shared in a more agile way. This means that artificial intelligence can accelerate medical decision-making in a pandemic.

Can robots help in hospitals?

A robot has been implemented for disinfection of hospitals and in everything that is labor, repetitive tasks in which it is possible to collaborate with humans, of assistance, can have a great role, disinfecting and attending patients who need a very close follow-up. Repetitive, the kind remember that you have to take this medication at this time, or that they come to scan that everything is fine, that there is no abnormality in the hospital wards. This is where issues of privacy, data protection, security of the robot itself come in, so it is difficult for us to see this implemented in hospitals in the short term. It would be something long-term, we would have to go ten years from now to see what happens with these types of robots, because within what are medical environments it is difficult to integrate them for security reasons.

“Technology has been geared towards the male figure of successful entrepreneur. Teaching your social impact can change it »


decoration

Are jobs at risk with the incorporation of robots?

Its implementation will be progressive, it will not happen in the short term. There is a high probability of applying this type of system in making decisions for specific and repetitive tasks. The digital skills of a fairly large sector of the population will have to be strengthened in positions of potential risk, but they will not disappear either. I don’t think they are going to destroy so much jobs because artificial intelligence needs a human to validate it, they are not going to replace them, but they will help us and we will supervise after pressing the button. In addition, other jobs will appear.

Will equality reach the classroom in science majors with an overwhelming majority of boys?

Regarding equality in the classrooms, there is still a lot to do, we must work on the gender perspective, something we are not used to and even less in stem careers. Traditionally, the entire technological issue has been oriented to the success of men, to the male figure of successful businessman, and the references and the telling of stories in another way, teaching the social impact of technology, it is true that it can promote this can change little by little. However, structural changes are needed based on temporary quotas and other measures that allow women to enter the decision-making bodies of high command, and that will be when everything begins to change downwards.


www.informacion.es

Leave a Reply

Your email address will not be published. Required fields are marked *