Politecnico di Milano (POLIMI) is a public scientific and technological university, founded in 1863, that trains engineers, architects, and industrial designers at Bachelor’s, Master’s, and PhD levels. It is internationally recognized for its strong focus on quality and innovation in both teaching and research, combining solid theoretical foundations with first-rate research infrastructures that enable a wide range of experimental activities. Research is a core mission of the university and is closely integrated with education, fostering high-level international results and continuous interaction between academia and society. Through long-standing cooperation and strategic alliances with the economic, industrial, and manufacturing sectors, Politecnico di Milano actively promotes technology transfer and applied research, while positioning itself as a reliable and recognizable point of reference for sustainable development in Italy and across Europe.
Paolo Zunino is Full Professor of Numeral Analysis and is a member of the Laboratory of Modeling and Scientific Computing within the Department of Mathematics. His recent research focuses on applied and numerical mathematics for partial differential equations; scientific computing and machine learning; computational biomechanics and oncology, with applications to digital twins and data-driven models for precision medicine.
He is the WP Leader and part of his team are Francesca Arceci, Piermario Vitullo and Cristina Macaluso, researchers of the Department of Mathematics.
POLIMI leads Work Package 7 (WP7) – Digital Twins, whose objective is to develop a hierarchy of digital twin models with increasing predictive capability to assess the risk of severe side effects in individual patients following radiotherapy treatment.
Main tasks are:
- to perform a retrospective study based on a historical cohort to develop low-level digital twins that provide a personalized radiation safety score after exposure to ionizing radiation;
- to develop a proof-of-concept for sex- specific digital twins;
- to develop high-fidelity digital twins for risk assessment leveraging on data collected prospectively (WP9), including a detailed baseline information and “environmental” data in the follow-up history;
- and to propose a personalized follow-up schedule guided by the digital twin’s prediction
To develop task-driven digital twin architectures capable of modeling toxicity risk at scale across heterogeneous patient populations, a Digital Twin (DT) framework based on probabilistic graphical models is adopted. This approach decomposes late toxicity risk into three distinct probabilistic components: exposure (e.g., radiotherapy dose, chemotherapy agents), susceptibility (e.g., age, imaging, biological, genomic, and transcriptomic factors), and severity (e.g., observed clinical outcomes). This decomposition provides a more transparent, flexible, and causally interpretable framework for understanding and quantifying toxicity risk.
POLIMI is also part of WP4 – Genetics, WP5 – Transcriptomics, WP6 – Software Tool and WP10 – Dissemination and Exploitation.
