Rose Collet

Background:
I graduated in 2020 with a BSc in maths at the University of Glasgow, where I also studied astrophysics for the first two years. My 4th year project focused on convection in a horizontal layer of fluid (Rayleigh-Benard convection).

Research Interests:

Keywords: Digital twins for healthcare, personalised medicine, patient-specific medicine, breast tumour modelling, soft tissue modelling, magnetic resonance imaging

My research involves developing and using patient-specific models, or "digital twins", to predict the outcome of neoadjuvant chemotherapy treatment for breast cancer.  My main goal is to extend existing digital twins to include the effects of fluid flow in the breast.

Digital twins are essentially “copies” of the patient’s tumour on a computer, which are built from MRI scans and evolved in time using mathematical modelling. The aim is to predict the outcome of a patient’s chemotherapy (hence this is also referred to as patient-specific or personalised medicine). We use MR images of the breast taken before and just after the start of treatment to calibrate the model to this particular patient. We can then predict how the tumour will evolve under the chemotherapy regime that the patient is undergoing. This is important information as there is no universally successful chemotherapy and the sooner a treatment regimen can be identified as unsuccessful, the sooner the patient can be moved to a different, potentially more successful treatment.

The model I use is based on the reaction-diffusion equation, which describes the spatio-temporal evolution of tumour cellularity in terms of tumour cell diffusion and proliferation, with a penalty term which represents cell death due to treatment. Tumour growth is restricted by accumulating stress in the surrounding tissue: this is represented by a “mechanical coupling” in which cell diffusion is damped by breast tissue stiffness. As chemotherapy is delivered intravenously, we anticipate that modelling the transport of drug through the blood should improve the accuracy of model predictions. We aim to do this using a poroelastic approach.

Why I chose the CDT in Fluid Dynamics:
I was keen to get a strong foundation in many areas fluid dynamics before starting a project, as my interests were originally quite broad. The first year was great for narrowing this down, and I ended up developing a strong interest in the medical applications of fluid dynamics which I had not previously been exposed to. The cohort structure also particularly appealed to me, as did the opportunity to meet supervisors before choosing a project.

Conference and seminar attendance/presentation

• Leeds Fluid Dynamics Symposium, Poster (June 2023)
• Computational Methods in Biomedical and Biomechanical Engineering Conference, Paris, Poster (May 2023)
• IACM Computational Fluids Conference, Cannes, Oral presentation (April 2023)
• Virtual Physiological Human Conference, Porto (September 2022)
• Research Computing at Leeds Conference, Oral presentation (July 2022)
• CISTIB PhD Group Seminars, Co-creator/organiser (2022)
• Leeds Fluid Dynamics Symposium, Organiser (June 2022)
• Leeds Fluid Dynamics Symposium, Poster (June 2021)

Student Education

• Co-supervisor for MSc Medical Imaging Research project, 2022-23
• Core Mathematics MATH1005 (2022-23, 2023-24)
• Computational Mathematics MATH2920 (2022-23, 2023-24)
• Introduction to Discrete Mathematics COMP1511 (2021-22, 2022-23, 2023-24)
• Fundamental Mathematical Concepts COMP1421 (2021-22, 2022-23)

Linked in Profile
Find out more about me on my Linked in account.