Medical student reflections - Matthew Shu and Kevin Hou 2024 essay prize winners
Tell us a bit about yourself?
Matthew majored in Data Science and Kevin majored in Computer Science in their undergraduate years. They were both in the guaranteed Medicine pathway and were brought together by their common computational and healthcare interests. Since beginning medical school, Kevin has been working in an AI imaging lab and a start-up called Resonait working on EEG biomarkers of depression. Matthew did a short stint at a healthcare start-up called Eucalyptus and concurrently participating in some exciting cardio-bioinformatics research on biomarkers of early resilience and susceptibility to coronary artery disease. They’re both deeply passionate about the good that this AI technology can bring to the world.
In your opinion, what is the most exciting development in medical technology or AI right now?
Autonomous agents is a fascinating new area of development. To put it intuitively, you have various individual instances of ChatGPT that all have its own memory and they work together to separate a large task into small pieces and complete them autonomously whilst coordinating together towards a final result. It’s already being used by developers around the world to deploy programs that would have otherwise taken months to develop. Can you imagine what an MDT would look like!
What did you learn about AI and technology during this process?
We have both been curious about AI since the beginning of university, and actively following the regulatory debate - a particularly memorable video including The A.I. Dilemma, from the Center for Humane Technology. We saw this process as the perfect opportunity to extend this debate to medicine; unpacking the technical infrastructure of transformers, the history of machine learning in medicine, and exploring both Australian and global regulatory legislation. Throughout the creation of the video we read through many relevant regulatory documents such as the EU Artificial Intelligence Act. It gave us a better sense of where the world is heading and how we should be positioning ourselves and contributing to ensuring patients and medical professionals can benefit from the new world order. Given the audience would mostly be medical professionals, we had the unique challenge of condensing just enough technical knowledge to help viewers break past a surface level understanding of the new era of transformer architecture and sequence models whilst still maintaining enough medical context to make it feel relevant and tactile. This exercise has not only improved our own intuition and understanding around the technical architecture itself but also developed our communication skills to be an effective and efficient bridge between these two worlds.
How do you see yourself contributing to the field of health or technology in the future?
We are passionate about both maximising outcomes from emerging technologies and minimising potential risk - and hope to blend our medical and technical backgrounds together! Matthew is interested in cardiology, aspiring to find novel methods of modelling cardiac risk and diagnosis in imaging, and translating this to patient outcomes. Kevin is curious about psychiatry, hoping to blend both implementation through and clinical research of emerging computational psychiatric methods. I think most importantly we both feel incredibly fulfilled when we work on projects that improves peoples' lives on a broad scale. We believe that we will create products that will enrich the doctor-patient relationship by being aware of what clinically matters but also the feasibility and potential of different technological approaches to solving important problems.