WP2 focuses on creating a semi‑automatic tool that makes it easier and faster for clinicians to evaluate PET‑CT images. A Positron Emission Tomography–Computed Tomography (PET‑CT) scan combines two techniques: a Computed Tomography (CT) scan, which produces detailed pictures of the body, and a Positron Emission Tomography (PET) scan, which shows how active different tissues are using a tiny amount of a radioactive tracer such as FDG (fluorodeoxyglucose). Together, these scans help doctors diagnose cancer, check whether it has spread, and assess conditions affecting the heart or brain.
The MONITOR‑PILOT and MONITOR‑RCT (Randomised Controlled Trial in WP3) studies use PET‑CT to assess the metabolic activity of lesions, based on the PREMIO criteria (developed in WP6), which build on the established PERCIST 1.0 (Positron Emission Tomography Response Criteria in Solid Tumors) framework.
Traditionally, quantification of metabolic activity is done manually. This process is complex and time‑consuming, requiring experts to carefully identify, outline (segment), and measure lesions over several time points to obtain a quantification which finally informs the application of the PREMIO decision criteria.
WP2 uses artificial intelligence (AI) to automate many of these quantification steps, making the process more efficient, objective, and consistent. The tool combines advanced image alignment, automated lesion tracking, and quantitative analysis into a single platform, reducing manual work while maintaining accuracy.
The project will deliver Digital PREMIO, implemented through RECOMIA, a secure cloud‑based platform designed to meet privacy and data‑protection requirements. The platform supports AI‑assisted analysis by automatically identifying VOIs (Volumes of Interest) and highlighting regions with high FDG uptake directly within PET and CT images. Importantly, clinicians stay in full control. They can always compare AI-generated quantifications with their visual inspection of the images and review and adjust the AI-led process as needed to support decision‑making and treatment planning.
Designed to work with both standard and modified PERCIST protocols—and openly accessible to the research community—this tool will support future multicentre trials and help advance more precise cancer monitoring. It will be available to all participating sites and will also support AI algorithm validation activities in WP7.