SPOKE 2
Smart Health and Data
Data Science for Healthcare 4.0: data management and the development of advanced methods, algorithms, and machine learning approaches that integrate large-scale clinical and molecular data.
Spoke Leader and Expertise
Alma Mater Studiorum Università di Bologna
The University of Bologna is internationally recognized for its expertise in data science applied to healthcare, advanced medical image analysis, and the development of artificial intelligence models to support clinical decision-making. Within the framework of precision medicine, Spoke 2 represents the digital infrastructure of the HEAL Italia project, enabling the integration, analysis, and interpretation of large volumes of clinical, biological, and instrumental data.
Partners
- BI-REX – Big Data Innovation & Research Excellence
- Engineering Ingegneria Informatica S.p.A.
- IRCCS IFO – Istituti Fisioterapici Ospitalieri
- Istituto Superiore di Sanità
- IRCCS Neuromed – Istituto Neurologico Mediterraneo
- Sapienza Università di Roma
- Università degli Studi di Roma Tor Vergata
- Università degli Studi di Cagliari
- Università degli Studi di Catania
- Università degli Studi di Milano-Bicocca
- Università degli Studi di Modena e Reggio Emilia
- Università di Pisa
- Università degli Studi di Verona
Scientific Coordinator
Prof. Stefano Diciotti
Mission
The mission of the thematic network (Spoke 2) is to transform healthcare data into clinically useful knowledge. In precision medicine, the availability of large amounts of data is not sufficient if this information is not integrated, analyzed, and interpreted in a reliable, transparent, and secure way.
Spoke 2 works to make data an operational tool for clinicians and researchers, supporting personalized clinical decisions based on quantitative evidence.
Objectives
- Improve the accuracy and timeliness of diagnosis.
- Support patient stratification based on individual risk.
- Make artificial intelligence a reliable and responsible tool.
- Promote personalized, data-driven clinical decision-making.
Activities
The activities of Spoke 2 cover the entire lifecycle of healthcare data. A first area focuses on the design of interoperable digital infrastructures capable of collecting and harmonizing data from different sources, such as electronic health records, imaging systems, diagnostic laboratories, and research studies. Interoperability enables different systems to communicate with each other, overcoming data fragmentation.
A second area of activity concerns the development of artificial intelligence and machine learning algorithms, that is, mathematical models capable of learning from data. In particular, the Spoke works on:
- Radiomics, the discipline that extracts quantitative information from radiological images
- Clinical decision support systems
- Explainable AI models, designed to be interpretable and understandable
The activities also include the definition of data quality standards, the management of security and privacy, and the validation of the developed models in real clinical settings.
Areas of work
Integration of confidential clinical data with omics and imaging maps
- Development of a collaborative platform for the integration of molecular, clinical, and imaging data
- Swarm Learning for the decentralized and integrated processing of biomedical data
- Prototyping of Software as a Medical Device (SaMD) for precision medicine
Evolution of Artificial Intelligence (Climbing AI)
- Beyond supervised learning: new frontiers of Machine Learning
- Advanced integration and modeling of multi-omics data
- Digital Twins for computational modeling and personalized intervention
From new methodologies to clinical applications: AI for personalized medicine
- Design of Artificial Intelligence techniques for augmented reality in robotic surgery
- Development of Network Analysis algorithms for the study of complex systems
- Development of AI-based medical imaging systems for diagnostic support and radioprotection in chest CT scans
Projects Funded through Cascade Calls
ARISE
Ai-based medical swaRm learnIng prototype for SEcurity and analysis optimization on multicentric clinical data
Develop an innovative prototype of a hardware-software platform based on swarm learning, using blockchain and artificial intelligence to optimize the processing and security of multicenter clinical data. The goal is to improve diagnostic accuracy and the personalization of treatments in precision medicine through: (1) OBJ1: Development of the ARISE middleware, (2) OBJ2: Development of AI/ML applications for precision medicine, (3) OBJ3: Integration and demonstration of the ARISE technology in a tri-institutional network consisting of three nodes at the Universities of Rome, Verona, and Bologna.
VET S.r.l.
BISTOURY
3D-guided roBotIc Surgery based on advanced navigaTiOn systems and aUgmented viRtual realitY
The BISTOURY project proposes an integrated system that combines MIRS, Augmented Virtual Reality (AVR), and an innovative artificial intelligence pipeline to provide surgeons with enhanced perception of anatomical structures through the integration of visual instructions and haptic feedback. By merging intraoperative decision support with advanced anatomical navigation guidance, the platform transforms complex data streams into real-time digital assistance, optimizing surgical accuracy and patient safety through an immersive and precise sensory interface.
PRIME
Private and Inclusive Synthetic Medical data for precision medicine powered by Swarm learning
The PRISMS project aims to transform the healthcare data ecosystem through the pioneering generation of multimodal synthetic data and the advancement of decentralized learning systems. The initiative places a strong emphasis on ensuring privacy, fairness, and bias mitigation, while promoting high standards of quality and trust in AI-generated data. Through synergistic collaboration with healthcare stakeholders, the project seeks to maximize the impact and scalability of its innovations, ensuring that technological progress translates into concrete, reliable, and secure real-world solutions.
