Rachel Thomson

Fellow in Clinical AI, Cohort 3

Fellowship Bio

I am a final year public health registrar working in Scotland, currently based between Public Health Scotland, NHS Ayrshire & Arran, and the University of Glasgow. My interests include policy evaluation, policy modelling, health inequalities, and environmental sustainability.

Fellowship Project

Evaluating environmental sustainability in deployment of AI-enabled clinical pathways in radiology
NHS Greater Glasgow and Clyde

This project was based with the Digital Health Validation Lab in Glasgow, and focused on the assessment of environmental sustainability outcomes (like carbon emissions or energy use) alongside clinical outcomes in evaluations of AI technologies in healthcare. This is important because AI interventions typically have higher energy and data demands than more traditional digital interventions, and concerns have been raised that this is not currently being sufficiently taken into account when deploying AI in clinical pathways.

A systematic review of existing evaluations of AI technologies in radiology found very few that integrated any consideration or discussion of environmental sustainability, with even fewer trying to measure these impacts. No studies existed which compared the carbon costs of an AI-enabled pathway versus standard-of-care i.e., treatment as usual, which makes it difficult to judge whether any clinical benefits of the technologies could be outweighed by their environmental impacts. Using guidance from the Sustainable Healthcare Coalition, a carbon calculator was devised to be used on data from the RADICAL (Radiograph Accelerated Detection and Identification of Cancer in the Lung) trial in NHS Greater Glasgow and Clyde (GGC). This will address the research gap identified in the review by directly comparing the environmental costs and benefits of NHS GGC’s original urgent suspicion of cancer clinical pathway for chest x-rays with a new AI-enhanced pathway, which introduces an immediate AI review of the x-rays to achieve faster prioritisation and follow-up for patients judged most likely to have a lung cancer. Work on this element of the project remains ongoing.

Fellowship Testimonial

For me, the fellowship has meant a gift of dedicated time to focus on developing my understanding and practical experience of one of the key technologies that will undoubtedly influence my future clinical career as a public health consultant. I have found it hugely valuable, with considerable flexibility built in to prioritise the learning needs which are of most importance to each individual fellow. In my case, I was able to focus on a topic which has huge public health importance – environmental sustainability – while still rooting this alongside technical and practical upskilling on AI integration into clinical pathways.

I particularly enjoyed the many training bootcamps across different UK locations, as it was fantastic to be able to learn from genuine leaders in the AI field as well as hearing what other fellows were working on. The huge variety in topics, specialties, and projects meant that it was always really stimulating and exciting to hear the cohort talk about their progress and development. I also had a great experience at the DHVL working directly alongside their Clinical Innovation Fellows, and feel like I have developed a fantastic network of like-minded colleagues who I know I will work with again in future. Overall the year has been hugely enriching and left me with so much learning which I will take forward into my future career – I wouldn’t hesitate to recommend it to others in public health or any other specialty!