Breaking into AI: Insights from a resident doctor
Article by: Dr Katherine Mackay
Dr Katherine Mackay, ST7 Clinical Oncology at Charing Cross Hospital, shares her advice on getting started in clinical AI and what the future of oncology training might look like.
What sparked your interest in AI?
AI is going to have an enormous impact on so many aspects of our lives. Healthcare, and the delivery of cancer care in particular, is at the forefront of this revolution. I was fortunate to complete my MD at the Institute of Cancer Research looking at the development and evaluation of AI auto-contouring systems for cervical cancer radiotherapy. After this, I had the opportunity to continue and expand this work as an NHS fellow in clinical AI. It’s really exciting to be working in a field with such huge potential for transformation. I’ve found that the more work you do, the more you realise there is to do, and my interest has just continued to grow.
You’re presenting at the RCR’s Global AI Conference 2026 – can you tell us about your session and what residents will gain from attending?
I’m lucky to be presenting in two sessions. One session is with other NHS Clinical AI Fellowship alumni, where we’ll be talking about the fellowship programme and other opportunities available for residents interested in AI. I hope this session will be useful for residents across all specialties who are considering applying for AI fellowships or who want to get involved in AI.
The other session is a debate on whether residents will need to learn to contour manually in the future, given developments in auto-contouring. This session includes representatives from the RCR CO AI Advisory Committee and the ORF. The debate should be really exciting and informative for clinical oncologists. AI is going to transform our practice, with huge implications for how we train the next generation of oncologists. It’s important that oncologists are proactive, rather than reactive, in changing how we work in the AI era. We need to decide how we are best going to harness the opportunities and meet the challenges of AI in radiotherapy planning. I hope this debate will be interactive and spark discussion. As this topic will affect residents, it’s crucial that we are part of the conversation – so please come along, get involved and have your voice heard.
What skills or perspectives have you gained so far that you didn’t expect when you started?
I quickly learned not to underestimate the importance of involving clinicians in AI evaluation and implementation projects. The development and implementation of AI for clinical use requires a huge range of varied and complementary multidisciplinary expertise, including clinicians, engineers, physicists, legal and management experts and computer scientists. When evaluating whether AI technologies have a clinically meaningful benefit in the real world, the role of a clinician is key. We can provide valuable insights about clinically utility and acceptability. Ensuring clinicians are involved helps us develop technologies that address areas of greatest clinical need.
What advice would you give to residents who are curious about AI but aren’t sure where to start?
It can be hard to take those first steps, but there are so many opportunities out there. Don’t be afraid to put yourself forward. AI is currently having its ‘moment’, so there are loads of options, such as attending online lectures or courses, getting involved with your local digital teams, applying for research projects or fellowships and of course, coming to the RCR Global AI Conference!
I’d also say to anyone starting an AI project, not to underestimate the importance of clearly defining the clinical problem before you start. Having real clarity about what you are trying to achieve with your AI makes it much easier to ensure there is a clinical need, define how you are going to evaluate it and embed it in your clinical workflow. The most useful AI technologies will be those that actually improve workflow efficiency and patient care, so it’s best to engage with an initial step of identifying workflow bottlenecks or areas that need improvement.
Where do you see AI having the most meaningful impact on oncology and patient care in the next few years?
AI is likely to impact the whole oncology workflow, from diagnostics (such as imaging and histopathology reporting) through to treatment decision-making (including MDTs), prediction modelling and risk stratification for follow-up. It also has huge potential for administrative support such as ambient scribes and clinic scheduling support. We’ve already made so much progress with auto-contouring for organs at risk and organ/anatomy-based target structures in radiotherapy planning. In future, these systems are likely to contour gross tumour volumes and consider uncertainties with more complex target structures. AI-based auto-planning will also continue to expand. These technologies will hopefully facilitate moves to fully online adaptive workflows.
One of the most exciting recent developments in auto-contouring is the ARCHERY trial, which is one of the first, large-scale prospective international AI radiotherapy trials ever done. This international, multi-centre prospective trial is investigating the quality and economic benefits of using AI-based, automated, radiotherapy treatment planning in cervical, head and neck and prostate cancers. The early results, presented recently at ESTRO 2026, are very exciting, and the software used has huge potential to improve access to curative radiotherapy in low- and middle-income countries. This could be a major step towards improving equity in cancer care and outcomes in areas of greatest need.