Challenges of the use of Artificial Intelligence in Radiology: Experience of the X-COV project.

By Juan Martinez - Marketing Director, Radiologia S.A

1 September 2021

The 7th SEFM and SEPR joint congress was held virtually from May 31 to June 4, 2021.

During the virtual congress, we were fortunate to attend the technical session given by Mr. Joaquín López Arráiz, professor of the Nuclear Physics Group and IPARCOS of the Complutense University of Madrid, and collaborator of the Health Research Institute of the Hospital Clínico San Carlos (IdISSC), Madrid. Being one of the experts in the field of Medical Physics in our country. Professor López gave a session on the Challenges of Artificial Intelligence and his X-COV project.

In the first part of the session Professor Lopez introduced us to the trends in the field of Artificial Intelligence (AI) in Medicine, last year there were 19,000 PUBMED publications on this topic. Another interesting fact was the AI market estimates for the next five years, reaching a figure of 30,000 million dollars in 2025, which represents a growth of 44% per year (source Markets and Markets Reports).

After putting into context the relevance and growth that AI has had, Professor Lopez, presented the birth of the X-COV project, in 2020 when the Pandemic started, doctors had to make many quick and critical decisions about patients with COVID-19, and answer the following questions: Should we hospitalize them? Is the treatment working? Are they ready to be discharged? And to this end, we started to think about training the algorithm to diagnose pneumonia.

Professor Lopez, explained the challenges in the field of AI, what are the approaches to work to develop from the regulatory point of view, how we can explain and understand the results, the reliability and robustness of the model, how we must classify the data to train the algorithm, the importance of working in multidisciplinary teams, to have a holistic approach when facing the problem, the radiologist is a key figure when facing AI models, as they are the ones who help us to select and interpret the cases. Concluding that from an ethical point of view, we should work with data balanced in gender, age, race, etc. To avoid sample bias.

If you want to know more in detail about the project, we recommend you to visit the website of the Complutense University.

Related Articles

Good Practice Guidelines for Technology Renewal and Update in Radiology (SERAM)

Challenges of the use of Artificial Intelligence in Radiology: Experience of the X-COV project.

Temporary emergency Hospital of Madrid region.