ESTRO 2022 - Abstract Book

S467

Abstract book

ESTRO 2022

to the need for radio-translucency and avoidance of irradiating through radiation-sensitive components, standard receiver arrays have a smaller number of individual coil elements and advanced RF technologies such as parallel transmit or higher order shimming might not be available. Furthermore, imaging gradients coils on the first clinical MR-Linac systems trade performance for spatial accuracy, resulting in lower peak amplitudes. The latter poses challenges in fast imaging and diffusion-weighted MRI in particular. Despite these challenges a plethora of qMRI techniques were demonstrated feasible on clinical MR-Linac systems, including chemical exchange saturation transfer (CEST) imaging, MR-relaxometry including MR fingerprinting and diffusion-weighted imaging. For some of these techniques a particular aim was to ensure reproducibility of the quantitative measurements acquired on normal clinical scanners. The consensus guidelines for diffusion-weighted imaging on the Unity MR-Linac system are an exemplary consortium effort, where the matching of diffusion time, a parameter that is not on every MRI scanner directly accessible in the user interface, was identified as more crucial than the matching of b-values. Reproducibility of qMRI measurements on dedicated MRI scanners is an important consideration as follow-up examinations of patients after completion of radiotherapy are unlikely to be carried out on MR-Linac systems. While qMRI faces challenges for integration on MR-Linac systems, it holds potential to gain insights into radiobiology, as measurements can be directly related to the delivered radiation dose. qMRI provides a unique opportunity to study hypofractionated or dose-escalated treatments, which become possible due to the reduced uncertainty resulting from daily adaptation on MR-Linacs. Due to the limited number of different MR-Linac systems clinically available, the fact that the same scanner platform is available to a large number of institutes is a stimulus for multi-centre clinical trials, in particular for the treatment of rarer cancers. Working on the same platform further supports and encourages the use of consensus protocols and analysis tools for qMRI.

SP-0537 The future of qMRI in radiotherapy

C. Fuller

1 USA Symposium: RTTs: AI and digital awareness

SP-0540 AI is an essential component of a curriculum for RTTs

M. Coffey Ireland Abstract not available

SP-0541 The changing face of the RTT profession in the AI environment

C. Gillan 1 , B. Hodges 2 , D. Wiljer 2 , M. Dobrow 3

1 University Health Network, Joint Department of Medical Imaging, Toronto, Canada; 2 University Health Network, University Health Network, Toronto, Canada; 3 University of Toronto, Institute of Health Policy, Management & Evaluation, Toronto, Canada Abstract Text Artificial intelligence (AI) in healthcare will require consideration of employment, training, education, and professional regulation. Recognizing that we are best served by taking a proactive approach to considering the nature and potential scope of AI, both in its benefit to our patients and in its impact on healthcare and those who practice it, it is important that we equip ourselves to engage in the relevant conversations. In radiation medicine we find ourselves at a crossroads, where our professions need to decide how they envision the impact of AI in our practice, and how we can collaboratively define appropriate AI-enabled care alongside society, industry, and other stakeholders. Gains in quality and efficiency in radiation medicine practice will require new workflows, skills and even models of care for all relevant professional groups. As we work to separate the reality from the hype, the cautious optimism from the fearmongering, and the human opportunities from the expansion of technology, we can begin to prepare for the future. Doing so will require an acknowledgement that we are not simply replacing humans with AI within the existing model of radiation medicine practice, but rather fundamentally disrupting practice by augmenting human abilities. This session will highlight the professional, ethical, regulative, and educational considerations around AI that should be as equally emphasized as the clinical and technical advancements in order to ensure responsible integration while maximizing potential. Technology is only as good as the people and system equipped to support it.

SP-0542 Dealing with a national cyber attack: A multi-disciplinary approach

A. Flavin

1 Ireland Abstract not received

Mini-Oral: 13: Implementation of new technology