Partner presentation: CNR – National Research Council of Italy (CNR-ISTI and CNR-IFAC)
The National Research Council (CNR) is the main public research organisation in Italy with the aim of carrying out, promoting, disseminating, transferring and improving research activities in the main fields of knowledge. The CNR consists of around 100 different research institutes, ranging from human and social sciences to engineering. In ProCAncer-I, the two CNR units involved belong respectively to:
1. “Alessandro Faedo” Institute of Information Science and Technologies (ISTI), Pisa
2. “Nello Carrara” Institute of Applied Physics (IFAC), Sesto Fiorentino (Florence).
ISTI-CNR is committed to producing scientific excellence and to playing an active role in technology transfer. The field of competence covers information, science, related technologies and a wide range of applications. The ISTI CNR unit is led by the “Signals & Images” Lab (https://si.isti.cnr.it/ ), a research group working in the elds of computer vision, signal acquisition and processing, image understanding, artiffcial intelligence, knowledge representation and modelling, intelligent vision systems and multimedia data understanding. In IFAC-CNR, the main research lines pertain to the general fields of optoelectronics, spectroscopy, and ICT, including the investigation of novel applications in several branches of interdisciplinary sciences, such as photonic devices and data analysis techniques for healthcare and well-being, mainly carried out by the BioPhotonics and Nanomedicine lab https://bpnlab.i-fac.cnr.it/artificial-intelligence-for-health/).
IFAC-CNR’s expertise also includes complementary aspects related to healthcare, such as privacy, ethical/legal issues and clinical policy. Both units are involved in different projects aiming to build imaging biobanks (the first for Tuscany region) and platforms, to boost 4P precision medicine in oncology by advancing translational research based on quantitative imaging and multi-omics analyses, towards a better understanding of cancer biology, cancer care, and, more generally, cancer risk. The expertise of ISTI-CNR and IFAC-CNR in radiomics and deep learning is fundamental both for the devel opment of AI validated models for the clinical scenarios envisaged in ProCAncer-I, and for ensuring that such AI tools and models comply with the FUTURE AI philosophy towards Fair, Universal, Traceable, Usable, Robust and Explainable AI.
In this respect, CNR is currently involved in the activities of WP4 (image repositories, sharing mechanisms, annotation, curation and standardisation methodologies), WP5 (development of master models), WP6 (development of vendor-specific and vendor-neutral AI models) and WP7 (clinical evaluation of AI models), and is leading Task 6.5: “Explainable, interpretable, ethical and trustworthy AI framework”. In the latter task, CNR’s activities are mainly dedicated to improving the trustworthiness of the solutions to be delivered in terms of (i) fairness and privacy, (ii) security and robustness, (iii) explicability and interpretability, and (iv) reproducibility and verifiability. Such an investigation will be carried out both by analysing the state-of-the-art literature and by fostering close collaboration between developers (e.g. computer scientists and engineers) and domain experts/end users (e.g. clinicians) in order to agree on the specifications of the ProCAncer-I solution and to clearly define a set of performance metrics to assess the value/efficiency of the AI framework in terms of trustworthiness.
Finally, with regard to the system transparency of the AI models, the CNR-ISTI is working on the definition and development of the traceability framework of the AI-based solutions and models, based on novel concept of the AI- Passport, jointly designed with FORTH, and on the definition of specific performance metrics and evaluation criteria to be used for monitoring the performance of the models after deploymen
