Workshop Organizer: Professor Dr John Damilakis, IOMP President
Lecturers: John Damilakis, M. Mahesh, Maryam Al Hashim, Shady Alkhazzam
The workshop titled “Advances in CT Dosimetry and Machine Learning: Optimizing Patient Radiation Safety” was organized during the MEFOMP 2025 Medical Physics Conference https://mefomp-conference.com/workshop_1/ This workshop spanned four days, from February 7 to 10. It was structured into both theoretical and practical sessions to provide participants with an understanding of contemporary issues in CT dosimetry and the integration of machine learning techniques to enhance patient safety.
Theoretical sessions focused on the evolution of CT dose metrics, emphasizing the transition towards personalized dosimetry approaches, dosimetric considerations of CBCT, an introduction to the workflow of machine learning applications in medical imaging, covering data acquisition, preprocessing, model training, and validation, a critical discussion on the inherent biases in AI algorithms, guidance on creating radiomics signatures, and instructions on transitioning from traditional Excel-based analysis to advanced big data platforms for creating interactive DRL dashboards.
A practical session provided exercises focused on developing regression and classification machine learning models pertinent to medical image analysis. The new AI-driven tool iDose designed to enhance accuracy and efficiency in CT dosimetry was introduced. Furthermore, participants applied their theoretical knowledge to develop radiomics signatures on CT images, reinforcing learning through practical application. Attendees also built interactive DRL dashboards using big data tools, with a focus on advanced features and customization.
The last day, a comprehensive overview of the ethical considerations, regulatory frameworks, and legal implications associated with deploying machine learning in medical settings was presented.
