AI applied to research for prevention, diagnosis, and treatment

It was essential to establish a common language between technical and medical experts to collaborate effectively with the various stakeholders involved in the project. Through the adopted methodology, several challenges were collaboratively addressed:

• enabling centralised data governance to ensure efficient access and use of information throughout the patient’s journey;
• identifying and controlling data sources to leverage predictive medicine, even when managed by third parties;
• adopting a more horizontal patient care management approach to optimise resources and create efficiencies;
• promoting value-based healthcare by prioritising treatments that uphold high standards of care.

One-on-one interviews were conducted with doctors and patients at the center to reconstruct the journeys related to different oncological pathologies. The cases were chosen to represent diverse hospital treatments and include a wide range of patient management data. Lung cancer treatment involves a combination of pharmacological management and surgical interventions, while leukemia requires radiological and pharmacological treatments.

This complexity necessitated analyzing heterogeneous data, such as demographic, laboratory diagnostic, and medical imaging information.

The adoption of AI supports medical researchers and enhances their diagnostic capabilities, especially in interpreting diagnostic images and analyzing complex data. Large language models (LLMs) scan medical records to identify trends, deteriorations, or adverse drug reactions. Trained on data collected by pathology experts, and based on deep learning systems on neural networks, the AI system improves performance across the medical team, offering valuable diagnostic comparisons that complement the work of clinicians, while leaving final decision-making in their hands.

In traditional management, patients are often the only ones with a comprehensive view of their clinical situation. This is particularly true for those with complex conditions, such as oncology or multisystem diseases, who are treated in separate departments using non-communicating data management systems. In the new scenario, advanced technology consolidates data from all departments, providing an integrated view of the patient, a critical advantage for multidisciplinary teams in Cancer Centers.

Remote monitoring via IoT enables automatic collection of vital signs at home, reducing the need for manual diaries and alerting doctors in the case of critical values. To enhance the doctor-patient relationship, we designed in detail a system that includes interactive tools, providing exam information and suggesting appointments, freeing doctors from direct computer-mediated interactions.