The development and implementation of different technologies in favor of the health area is already a recurring practice in the daily life of the health area, but these advances, developments and technological implementations have not been entirely under a regulatory framework, a product of the development exponential technological growth seen in recent years. At this point, Artificial Intelligence (AI), used in clinical trials in recent years, should not be subject to audits and accountability regarding its design and development. Which allowed its capabilities to be exaggerated, but the implementation of new standards may be changing this reality.

An international consortium led by a group of medical experts has presented the first official standards for clinical trials mediated by Artificial Intelligence. This results in a set of manipulated or unverified information that magnifies the results of AI and that can affect the credibility of this technology.

The new standards show a set of rules on how clinical trials should be carried out and how they should be socialized, these rules are already in practice around the world for the development of medicines, diagnostic tests and other medical interventions. Now researchers will have to show the skills necessary to use this technology, its evaluation environment, details of how to interact with it, analysis of error cases and other important points.

A standardization of this technology will help us understand in depth and clearly the effectiveness and safety of a treatment or a clinical technique. This will allow us to make comparisons, analyzes and evaluations of various applications of Artificial Intelligence in the sector with scientific rigor, avoiding the bad practice of some developers who boasted or inflated their results without having to face the validation and control that is being done. applying to other types of interventions or diagnoses. A study in the BMJ has well warned this, indicating that offering information without proper research, or exaggerated claims, could represent a risk for millions of patients.

For now, it can be known that in the very near future these new standards will facilitate an objective, efficient and independent assessment of all uses and applications of this technology that will lead to better and more reliable medical AI.

To learn more about this topic, we invite you to enter through the following link:

Heaven, W. (2020) “These international standards will stop AI from fooling doctors” MIT Technology Review. Retrieved on 12/04/2020 from: https://www.technologyreview.es/s/12631/estas-normas-internacionales-impediran-que-la-ia-engane-los-medicos