AI education intervention for medical imaging professionals: development, delivery and outcomes

Background: Artificial intelligence (AI) has become incipient in medical imaging and is significantly impacting the field. Despite the rapid implementation of AI tools to clinical practice, there remains a paucity of research to investigate the educational needs of staff who are expected to effectively utilise this technology. Earlier stages of the REAL_AI project gathered data on these needs and requirements, which was used to curate an educational intervention aimed at enhancing the AI awareness of medical imaging staff. AI readiness has been identified as a precursor to AI adoption, and education is a tool to combat fear and resistance which is a barrier to adoption. Methods: the team were approached by Health Education England’s E-Learning for Health (HEE ELFH) and tasked with designing an e-learning module tailored to medical imaging staff in the UK. The multidisciplinary research team at Ulster University used data from the initial phases of the REAL_AI study to curate a beginner module on AI. The intervention was hosted on the ELFH learning platform, to a potential audience of over 70,000 users. To assess impact and evaluate overall effectiveness of the intervention, short pre- and post- surveys were designed to capture basic demographics and identify participants’ awareness level. ELFH also
captured usage data and engagement metrics. Surveys included a mixture of Likert scale, multiple choice and open-ended questions. Data are being collected by ELFH and will be anonymised prior to being returned. Analysis will be performed using SPSS and N-VIVO.
Results: Data analysis is scheduled to begin in May 2024 and will be completed for presentation at the EuSoMII conference in October 2024. Participant feedback will determine the impact of the intervention and allow for evaluation of the tool. Metrics assessed will include any change in awareness post-intervention, and engagement and completion rates. This will help inform future iterations of the e-learning module and could inform the adaptation of the curricula in medical imaging. Conclusion: It is hoped that the findings of this study will provide a vignette of the current awareness of AI and insight as to the impact of this targeted method of delivery, the content of the module, and lay a foundation for scaling and disseminating similar interventions across the wider healthcare community.