A team of medical students at the University of Oxford has recently seen their research published in a peer-reviewed journal, as part of their training with the Centre for Evidence-Based Medicine (CEBM). The work, originally developed during the CEBM’s Special Study Theme (SST) for undergraduates, demonstrates the real-world value of embedding evidence-based medicine into medical education.
A systematic review evaluated the diagnostic accuracy of artificial intelligence (AI) in radiology assessments and included 34 studies, of which 23 were included in the meta-analysis. The conclusion indicated that AI assistance tools could enhance clinicians' diagnostic performance in cancer diagnosis. Additionally, updated reporting guidelines may help address potential methodological limitations and clarify the value of AI in healthcare.
The students – working under the supervision of Jason Oke and Annette Plüddemann – conducted a systematic review according to rigorous standards, from formulating a research question and constructing a protocol to analysing and synthesising the evidence. This mirrors the pathway promoted by the CEBM, which, for decades, has championed systematic reviews and robust methodology to improve health care decision-making.
'Publishing in a recognised academic journal is a significant achievement for undergraduate students. It not only validates their hard work but also highlights the SST programme’s goal of providing students with both theoretical knowledge and practical experience in conducting high-quality research,' said Professor Carl Heneghan, Director of the CEBM. 'By doing this, these young researchers contribute to a broader evidence base that clinicians, policymakers, and patients can utilise. This illustrates the importance of equipping future doctors with strong evidence-based medicine skills from an early stage,' he added.
As the CEBM continues to promote student-led scholarship, its role in shaping the next generation of clinicians and researchers remains essential.
Dylan Zhao, Thomas Packer, Xiaobo Jie, Muhammad Shahid, Jason Oke, Annette Plüddemann, Diagnostic accuracy of artificial intelligence-assisted radiology assessment of cancer: a systematic review, BJR|Artificial Intelligence, Volume 2, Issue 1, January 2025, ubaf016, https://doi.org/10.1093/bjrai/ubaf016