ESTRO 2025 - Abstract Book

S2271

Interdisciplinary – Health economics & health services research

ESTRO 2025

Conclusion: While clinical research in oncology is experiencing unprecedented growth, radiotherapy does not seem to benefit from the same level of interest, delaying the development of technologies and therapies that could benefit patients. Furthermore, moreover, disparities in access to clinical trials remain significant on a global scale, underscoring the need for international collaborations to address patient needs.

Keywords: clinical trials, radiation oncology, inequality

References: Zhu H, Chua MLK, Chitapanarux I, et al. Global radiotherapy demands and corresponding radiotherapy-professional workforce requirements in 2022 and predicted to 2050: a population-based study. Lancet Glob Health 2024; published online Oct. DOI:10.1016/S2214-109X(24)00355-3.

1007

Digital Poster Development of an Electronic Framework to Guide the Procurement and Evaluation of Artificial Intelligence

Based Software in Oncology Mark Agius 1,2 , Susan Mercieca 3

1 Radiotherapy Department, Sir Anthony Mamo Oncology Hospital, Msida, Malta. 2 Department of Computer Science, Faculty of ICT, University of Malta, Msida, Malta. 3 Radiography Department, Faculty of Health Sciences, University of Malta, Msida, Malta Purpose/Objective: The integration of Artificial Intelligence (AI) in radiotherapy and oncology settings presents transformative potential for enhancing patient outcomes and operational efficiency. However, the adoption of AI solutions in clinical practice has been relatively low [1]. This study aimed to establish a comprehensive framework to guide the effective procurement and implementation of AI-based software for radiotherapy and oncology. Material/Methods: The study was conducted in 4 phases as shown in Fig.1. A literature search was first conducted to identify existing guidelines and key criteria for evaluating AI software. The findings were used to inform the semi-structured interviews. An expert panel consisting of 4 healthcare professionals (radiologist, radiographer, medical physicist, and clinical oncologist), 4 technical information technology experts, and 2 senior legal procurement officers were invited to participate in the interviews to identify the barriers hindering the implementation of AI software and the criteria that should be included in the AI evaluation procurement framework. The recorded interviews were transcribed verbatim and analysed through content analysis. The literature review and interview findings were used to develop the AI procurement framework. A questionnaire was then distributed to the same experts to assess the criteria of the AI framework. The participants were asked to rate each item within the framework using a Likert scale of 1 (exclude criteria) to 5 (essential) and to provide additional feedback about the framework. All criteria rated as 4 or 5 by at least 75% of the participants were included in the final framework.

Fig.1: Study workflow