Will artificial intelligence impact the world of healthcare?

An interview with Fabio Tedoldi, Head of Bracco Global R&D

Fabio Teboldi Head of Bracco Global R&D

Today, artificial intelligence and in particular machine learning algorithms helps developing more and more powerful tools to discover new drugs, to make production processes more reliable and to better manage patient care.


Alongside the growth of investment, in this area, a cultural change is needed to allow AI to spread widely and to consolidate in the clinical as well as in the industrial practice.


Fabio Tedoldi, Head of Bracco Global R&D, in this brief interview, illustrates how artificial intelligence and new machine learning techniques, can be useful for patient care and for the future of the pharmaceutical industry.

Watch the interview here

– or read the full interview below the video:

Please accept marketing-cookies to watch this video.

I see at least four areas where artificial intelligence, and in particular new machine learning techniques, can be useful, that are important already today but even more in the future of the pharmaceutical industry.

 

The first one is drug discover: in the pharmaceutical industry we always need to create new products, new drugs that are able to selectively recognize certain proteins and certain factors that are over-expressed in certain diseases. Today, artificial intelligence is starting to create more and more powerful tools to allow us to screen in vitro new products, new molecules and thus make the process faster and more effective.

 

Then there are specific products that we can develop, for example in radiology, which is my area within artificial intelligence, to improve image quality, speed up the workflow, extract more information than we are currently able to extract from images.

 

There is a whole industrial context in which, in manufacturing operations, we have quality data that, if properly interpreted, can give useful information to prevent certain production problems that can occur in the future.

 

Finally, there's a customer care topic where algorithms that are able to understand what are the customer needs, what are their preferences, what is the work rhythms, can give the pharmaceutical industry, a way to act better and more effectively.

 

So, thanks for the question, which is the reason I agreed to come although I’m not the maximum expert in the artificial intelligence field, but this was the message I wanted to communicate, that today we need Artificial Intelligence; in order that the specific techniques, especially data-driven techniques, becomes widely used, a cultural change is needed, a cultural change able to move away from software per se toward the understanding of how important is the data, the quality of the data, the availability of the data, the solution of a whole range of problems, particularly related to the medical field, like privacy, confidentiality aspects associated to this data, regulatory aspects.

 

These algorithms are artificial intelligence algorithms, therefore they can learn every day from what's going on, become more and more refined based on the everyday experience. But nowadays, the rules of the regulatory bodies do not allow to change software or algorithms every day, they require a certain approval process that is not yet ready and refined for continuously evolving algorithms.

 

And then at the end Let’s touch a topic about the implementation: so far, we have a certain environment especially in Radiology, which again is where I’m involved, in which there are certain players. If you want to make available any new ideas that can be interesting for the patient care, you need to create an environment that is so-called vendor agnostic and therefore can work on any platform that is today available in hospitals.

 

So I go back to the beginning, I think that a doctorate on this specific topic is very important especially to create a new generation of scientists that comes already with the right mindset to make these techniques, that are emerging today, widely used.