Gerard McCabe

Fellowship Bio

ST8 Oncoplastic Breast Surgeon based in Edinburgh, with a focus on reducing positive margins in breast cancer surgery. My research interests centre on applying convolutional neural networks to improve intra operative margin assessment and surgical outcomes.

Fellowship Project

This project explores whether artificial intelligence (AI) can help breast surgeons reduce the risk of leaving cancer behind during breast-conserving surgery. We are developing a high-quality database of specimen mammography images and using these to train a convolutional neural network (CNN) to identify imaging features associated with positive surgical margins. Ultimately, the aim is to provide surgeons with additional information during an operation, supporting decision-making and potentially reducing the need for further surgery. My contributions: I initiated and lead the project, securing funding as the primary applicant and building the multidisciplinary team needed to progress from concept towards clinical evaluation. This has involved coordinating clinical, AI, ethics, information governance and research expertise. I have worked closely with NHS Safe Haven teams and Caldicott representatives to establish secure access to patient imaging data, and engaged with industry partners to explore technical development and future clinical translation. Milestones: we have secured funding, established the project team, completed key ethics and information governance requirements, and developed the pathway for creating the imaging database. We are now starting our first trial within the secure Safe Haven environment, where the CNN will be developed and evaluated. Remaining work: next, we aim to demonstrate performance, secure further funding for a larger multi-site dataset, and progress to a shadow-mode trial with hospital PACS teams. This will test the AI alongside routine care, without influencing clinical decisions, before prospective clinical evaluation.

Fellowship Testimonial

Graduating from this fellowship has given me a much stronger understanding of how to approach innovation and AI research and how this differs from traditional medical academia. It has changed how I approach research and how to build a project from an initial idea towards clinical translation. At the start of the fellowship, one of the most daunting aspects of AI research was understanding the governance required to access and use healthcare data safely. The fellowship gave me the knowledge and confidence to navigate this through lectures, practical resources, professional networks and, particularly, via direct supervision and support from the Tayside team. This experience has provided me with a foundation that I can now apply to future AI projects. The aspect I enjoyed most was collaborating with the other AI fellows. Working alongside clinicians from different specialties, sharing ideas and learning from each other’s experiences has been invaluable. These relationships have already led to several new collaborations and projects with fellows. Looking ahead, I hope to use the skills, experience and networks developed through the fellowship to build a local collaborative AI network. My ambition is to bring together clinicians, researchers and technical expertise to identify and tackle meaningful problems in breast cancer care with a focus on developing AI solutions that can translate into genuine improvements for patients.