8 May 2024 – National Physical Laboratory (NPL), Teddington (UK)
The MAIBAI project celebrated its first 9 months of activity with the first Stakeholder Workshop. The workshop, held at NPL in the historical venue of Bushy House, was finalised to the standardisation processes needed for the management and categorisation of clinical data for AI application in the medical field. The agenda included the following contributions:
- Dr. Spencer Thomas (NPL): Welcome and Introduction
- Dr. Alessandra Manzin (INRiM): MAIBAI Project Overview
- Dr. Nadia Smith (Royal Surrey NHS Foundation Trust): Overview of OMI-DB, Categorisation of Data and Validation of AI
- Dr. Louise Wilkinson (Oxford University Hospital NHS Foundation Trust): Are we ready for AI? A Radiologist’s Perspective
- Dr. Rebecca Nutbrown (TÜV SÜD Nuclear Technologies Division): The Need for Regulation in AI-Driven Medical Imaging
Within the medical area, the attention was focused on breast screening, the exemplar investigated in MAIBAI, with Louise Wilkinson presenting the current approaches adopted in clinics, the opportunities for deploying AI in screening, and the need for monitoring at scale and for systematic evaluation. Within the standardisation area, useful feedback was provided by Rebecca Nutbrown about the current global regulatory framework, the aspects related to privacy, security, and ethics, as well as the need for reporting data sources, data set type, validation approach, and clinical utility assessment. The invited stakeholders also discussed about what is expected and what is understood from AI model applications in clinical diagnostics, as well as about the fundamental concept of AI explainability.