A key contributor to these accomplishments is Maggioli, leading efforts in designing the Radiomics AI Passport, a transparent tool for monitoring the training, validation and deployment of AI tools, including radiomic AI tools, by establishing the framework and technical design and developing the first-year prototype and user interface for the upcoming year.
The project has also successfully established criteria and metrics for a comprehensive radiomics AI evaluation in breast cancer and conducted explorative analyses of open-access datasets to enhance project capabilities. Additionally, RadioVal pioneered the creation of its dedicated clinical database and user manual, fostering standardized data collection and methodology to enhance the project's capabilities.
Furthermore, as an integral component of the social innovation framework, RadioVal undertook a stakeholder mapping initiative and facilitated a toolkit, involving 86 participants in co-creation sessions and interviews. It's crucial to note that this marks the initial phase of the ongoing social innovation activities set to span the entire project. Lastly, RadioVal diligently laid the groundwork for forthcoming clinical studies, ensuring ethical and regulatory compliance, and officially registered as a Clinical Trial Study (RADIOVAL NCT05070884).
Looking ahead to the second year, RadioVal is excited to announce visionary plans. The project will spearhead an international study for an in-depth evaluation of radiomics tools in breast cancer treatment planning, emphasizing criteria such as robustness, fairness, usability, explainability, traceability enabled by the AI passport that Maggioli is developing, and scalability. Consensus-building on multidisciplinary requirements and the development of evaluation metrics for Radiomics-based models and AI in healthcare are among the strategic priorities.
Furthermore, RadioVal will focus on the development of RadioNac, with an emphasis on automated breast tumor segmentation and the validation of breast cancer treatment prediction and survival models. A human-in-the-loop component is currently being designed by Maggioli and will be integrated with RadioNack, ensuring that humans remain in control throughout the use of AI tools. With these initiatives, RadioVal is committed to pushing the boundaries of innovation in breast cancer treatment prediction, bringing us closer to delivering personalized treatment on a global scale.