Federico Battiston, the Italian physicist who won the most coveted prize in network science

Everything is a network. Social networks, but not only. There are networks in the brain , in the villages of hunter-gatherers in the Philippines, in global economic systems, in the closed rooms of an escape room. Invisible networks that determine who we are and what we choose. And there is a science that studies them. It's called network science , or the science of networks, and it tries to understand how relationships between people, neurons, and institutions influence collective behavior .

«What matters in these systems, made up of very different components, are their interactions ». It is in these subtle architectures that Federico Battiston works, he deals with statistical physics, «that is, the physics of many objects», an Italian professor who has just been awarded the Erd?s–Rényi Prize, the most prestigious recognition in this field, named after two Hungarian mathematicians who invented the first network models at the end of the 1950s. A sort of small Fields medal for network science.

Battiston has given a strong boost to network theory, where connections are limited between pairs of individuals, such as in social networks. Instead, we need to look at groups, collective interactions, moments in which influences are not of one, but of many.

Thirty-six years old, associate professor at the Central European University in Vienna, Battiston directs the doctoral program in network science , the only one of its kind in Europe. He coordinates a research group of eight people, including doctoral students and postdocs, from all over the world: Indians, Chinese, Brazilians, South Africans and, of course, Italians.

Network science is an interdisciplinary field that deals with the study of all those systems that we can represent as networks .

« Not only social networks, but also ecological, biological systems such as the brain, transportation systems, economic exchanges or political alliances. All of them can be mapped as networks, that is, a collection of elements, the so-called nodes , and links that describe their interactions. Even if they are different from each other - people, for example, are different from elements such as neurons or nations - these systems also have many common properties linked precisely to their interaction structure. And it is precisely the study of their architecture that allows us to understand complex emergent behavior, as in the case of the spread of epidemics in a population, or an epileptic attack, which occurs when our brain synchronizes too much ».

Romano , classical high school, degree in theoretical physics at Sapienza. Two exams with Giorgio Parisi, Nobel Prize for Physics in the field of complex systems in 2021. Then he leaves: to London for a doctorate in mathematics with Vito Latora, one of the pioneers of network science. Then he moves on to neuroscience in Paris , anthropology in Zurich , three years in Budapest .

«Mobility is an almost necessary condition for making a career in academia. I think I managed to achieve good results because I had the opportunity to work in various laboratories. But moving is a difficult choice, which requires compromises, even strong ones, on a personal level» Since 2020 Battiston has been in Vienna .

Network science is very present in everyday life. And many concepts of this discipline have now entered common language. How many times have you heard of small world , the famous six handshakes that connect anyone, the six degrees of separation , or super spreaders that spread a virus faster than others?

"These are all concepts that come from network theory. Today , the algorithms that advise us what to read or who to follow on social media are based on network theory. However, improper use of these algorithms can lead to negative consequences such as the formation of echo chambers , online environments where we interact almost exclusively with people who think like us and we end up reinforcing and polarizing our prejudices."

In financial systems, the concept of too big to fail has now been replaced by too interconnected to fail : «The total exposure of a company or financial institution is not a sufficient parameter to capture the risk of a systemic crisis, but it is necessary to consider the intricate financial relationships between different institutions to understand how the failure of a bank can lead to a series of 'cascading failures'».

Thinking about Covid , network science has made fundamental contributions to epidemiology. «Using real data on social contacts and mobility, it is possible to obtain much more accurate predictions on the progress of epidemics compared to traditional approaches based on mathematical equations, which erroneously assume that all individuals are equally exposed to the risk of contagion, without considering their specific interactions. Finally, artificial intelligence algorithms produce increasingly effective drugs by identifying new combinations of active ingredients».

But network science sometimes also begins in an escape room .

"We collected data on interactions between people during the game and then observed who won and why. Who made decisions, who was central in the social network, who really contributed to the solution. A useful way to understand what makes a team effective, which interaction structures allow the emergence of cooperation and better performance."

The network also appears in the hunter-gatherer villages in the Philippines or Congo. Where there is nothing. There is no TV, telephone or Internet. «In these populations , knowledge, such as the use of certain plants for medicinal purposes, is passed down from family to family. Even our microbiome depends on our social interactions. In a recent study in Science , by linking the diffusion of cultural traits and genetic distance between different populations in the world, we have shown how chimpanzees also show signs of cultural accumulation that are due to the passage of knowledge between different generations, a characteristic that until recently was considered a prerogative of our species. Chimpanzees will never build a computer, but perhaps we humans are not so special after all...».

Network science is also applicable to climate change . "For example, we could raise awareness of climate change through collective action promoted by appropriate social interactions."

There is a young engineer who studies CT scan images to teach machines to read them better. She is only 25 years old, she is an expert in mathematical modeling, and she is convinced that there is much more clinical value in that data than we are able to exploit today. Her name is Angelica Iacovelli and she goes beyond research projects. She has developed algorithms with a specific goal: to automate CT scan analysis in the oncology field, a process that is still done manually in many hospitals and on which fundamental decisions depend. First of all: the type of treatment.

He studied at the Polytechnic University of Milan, did research at Stanford, founded a startup in the United States and today works between Milan and San Francisco. He collaborates with the oncologist Michele Ghidini of the Polyclinic of Milan and has a dream: "To ensure that research does not remain closed in papers, but becomes a real tool in the hands of doctors."

Hearing her speak is a marvel.

"We have developed software that automates the analysis of CT images to extract fundamental metrics that are not used today. Among these metrics, for example, we extract skeletal muscle mass. It is precisely from this metric that we determine whether a cancer patient is sarcopenic." He explains precisely that sarcopenia is the disease in which muscles and strength are lost and that according to the latest scientific studies it is correlated with a worsening of the outcomes of therapies and a higher mortality. "Today these analyses are performed manually or with tools designed for research and cannot be used in clinical practice (they have a very high margin of error). No one has yet introduced clinically operational software to identify sarcopenia directly from CT images. Our goal is precisely this: to bring this technology into the clinic, with results available in less than 2 seconds and without the need for the intervention of an external expert, such as a nutritionist."

Has anyone ever thought of this before? "In the medical field, innovation is very slow. You come up against regulations, politics, and long lead times . It's difficult, so often those who try give up. But someone has to do it. And if no one does it, I will."

Angelica works with mathematical models and artificial intelligence algorithms applied to the analysis of medical images. «I use, among other things, neural networks and segmentation techniques based on Hounsfield Units (density values ​​​​contained in CT scans), to automatically distinguish the different body tissues – such as muscles, visceral and subcutaneous fat – and derive clinically relevant metrics».

Her story begins in Bari. She grew up between Grumo Appula and Palo del Colle, two towns in the Bari hinterland. An only child, she came from humble origins. "My father was a janitor, my mother took care of the house. We never had many resources, we made great sacrifices to study."

After high school, she chooses Civil Engineering at the Polytechnic University of Milan. It is not a linear path: «At a certain point I found myself studying concrete and I realized that it was not what I wanted to do». She stops, reflects, changes. She enrolls in Mathematical Engineering: «I liked the idea of ​​acquiring solid tools, to then apply in different fields — biomedicine, finance, statistics». During her university years she wins several scholarships, goes abroad, a year in Spain, one in India. Then she graduates and the opportunity comes that changes everything: to do research at Stanford. «Even at that time I had to look for funds to be able to move. I started talking to everyone, writing everywhere, I applied for every scholarship possible, even those for which I did not have the requirements».

Eventually, she was selected by the Ermenegildo Zegna Founder's Scholarship, which supported her for a research period at Stanford. There, she worked on computational models to simulate the cardiovascular system, based on graphs and artificial intelligence. And she won the BHI award. "We developed a digital twin of blood flow in the cardiovascular system : an AI model capable of simulating the behavior of the cardiovascular system in a personalized way for each patient. I combined LSTM (Long Short-Term Memory) neural networks with Graph Neural Networks, obtaining improvements compared to previous methods. But the most impressive thing is the speed: traditional methods can require hours of calculation on supercomputers to simulate even just a single heartbeat, our model can do it in a few seconds."

An experience that made her understand many things. "I realized that we could build very powerful tools. But I wanted them to be really used and have an impact." So Iacovelli began to seek contacts, to talk to those who work in the startup world and received another scholarship from Des Traynor, co-founder of Intercom, a unicorn in Silicon Valley.

Then she joins Lead the Future , a network of Italian talents in the STEM world, designed to connect mentors and mentees, and she meets the oncologist Ghidini. «He spent hours telling me about the concrete problems that doctors encounter every day in the ward. He opened up a whole new world to me. I contacted hundreds of other oncologists, in Italy and abroad, to see if it was a shared problem. And I discovered that it was».

In the summer of 2024 Angelica founded Nucleo Research in Silicon Valley . The validation phase is starting.

The next steps in their work are crucial. "We want to bring into clinical practice a software to automate the measurement of tumor lesions and their classification into target and non-target, benign or malignant. Measurements are still performed manually in almost all hospital facilities."

In the meantime, Angelica is preparing the first round of fundraising with international investors. «I am evaluating some opportunities for support from international entities interested in the project».

It is also a personal issue that drives her. "I grew up in a context where the disease was present and I never had control over anything. My mother died a few years ago from an aggressive tumor. My father suffers from muscular dystrophy, today he is not conscious. If I cannot change the course of things, I decided to at least try to change the way in which they are faced."

Angelica also leads communications for iTAL Foundation , a non-profit organization that builds bridges between Italian talent and Silicon Valley.

His scientific vision is clear. «Today's detection models are already extremely advanced. I believe that the bottleneck is no longer in the technology. The problem is everything that comes after: regulatory barriers, data protection, long times, integration into clinical processes. The doctor must be able to trust disruptive tools. Our goal is not to replace it, but to give it an additional tool. And make it reliable».

What has really made a difference in your life?

«The desire to do something of my own. Often in the world of research or in that of large companies you are part of a cog. I want to have a direct, real impact, to see an idea transform into something that works. I am almost obsessed with my project. It comes first in my life. And if I want to increase the probability of making it I have to be in Silicon Valley. When you are in San Francisco and you go to an event, no one asks you: How are you? The first question you hear is: what are you building? There is more: in Italy you always feel like a girl “too young” and they don’t give you credibility. If you say: I want to cure cancer, they answer you with irony or with: “What are you saying?”. If in California I say the same sentence the response is: “Great! How do you plan to do it? Can I help you?”.

We young people can inspire you. We can change the world, we can do it even without following traditional paths. We need confidence. This is why I tell my peers not to be discouraged by pessimism. Which is nothing more than a defensive approach, a protective shield, a way to avoid taking risks. But the future is not built by staying safe».

Battiston has worked in many places in his career. But never in America. «I spent several months in high school in the US and travelled a lot afterwards. But I'm not a huge fan of the American research system, which is based more on competition , understood in the Darwinian sense of survival of the fittest, than on cooperation. In Europe we have fewer resources, but in our craftsmanship we are good at doing well with the little available. The Erdos-Renyi prize is almost always won by researchers who are American or who live in the United States. And today I consider this prize as recognition of our European community, which has made fundamental contributions to network science. And on Italian universities he adds: «The same critical issues of Italy highlighted in the newspapers are often found abroad, sweetened by increased funding. The money from the PNRR has helped relaunch our country, but now we need structural funds to consolidate what has been achieved».

Would you return to Italy?

"If interesting proposals were to arrive, I would listen carefully. Currently, my relationships with the Italian are in the sports field. I am part of the national masters of the over 33s of Ultimate Frisbee, a beautiful sport that I recommend to anyone who wants to try something different". And then he adds: "Each of us should find a field that is passionate about them. And spend time with experts, supervisors and mentors. Because, as the science of networks teaches, the diffusion of knowledge requires interaction ".