We are pleased to announce the next MiLS meeting titled "From bioimages to mathematical models", which will take place at University of Exeter (Streatham Campus) on 24th June. This meeting will explore the state-of-the-art in UK research focused on using modern techniques to bridge the gap between biological imaging and mathematical modelling. The meeting aim is to provide an updated vision of the pipeline used to transform bioimages into robust inputs for mathematical frameworks.
Travel Information
The meeting will be held at University of Exeter (Streatham Campus) in Margaret Rooms 2&3 in the Queens Building (no. 11 on this campus map). Information on how to get to the venue, including public transport options, cycling facilities, bicycle routes and parking details, can be found here.
Registration
Attendance to the meeting is free of charge, but we kindly ask you to register your intention to attend by completing the registration form. Registrations and contributions of posters will be accepted up until 17th May December and can be made through the registration form.
We look forward to welcoming you!
PROGRAMME (abstracts can be found below)
10:00–10:15 Arrival, registration & coffee
10:15–11:05 Dr Remy Chait (University of Exeter)
"Coexistence and transport of lytic phage infections with migrating bacterial hosts"
11:05–11:55 Prof. Matthias Ehrhardt (University of Bath)
"Getting the most from data: Modern mathematical approaches to inverse imaging problems"
11:55–12:45 Dr Peyman Shadmani (University of Exeter)
"Image analysis and computer vision to inform mathematical models in cell biology"
12:45–14:00 Lunch & poster session
14:00–14:50 Prof. Tanniemola Liverpool (University of Bristol)
The mathematics and physics of wound healing
14:50–15:05 Coffee break
15:05–15:55 Prof. Kirsty Wan (University of Exeter)
Imaging motion at the microscale
15:55–16:45 Dr Joshua Bull (University of Oxford)
"From imaging to interactions: spatial methods for parameterising agent-based models"
16:45 Close and pub trip
ABSTRACTS
Speaker: Remy Chait (University of Exeter)
Title: Coexistence and transport of lytic phage infections with migrating bacterial hosts
Abstract: Phages, viruses that infect bacteria, coexist alongside their hosts across marine, terrestrial, and animal environments. Yet the infection dynamics of virulent phages appear to disfavour stable coexistence. Invading phages rapidly kill bacterial populations in well-mixed in vitro environments, eliminating hosts outright or setting off rounds of selection for phage-resistant bacterial mutants, followed by host-range expansion phage mutants, ultimately culminating in collapse of phage populations. This gap in outcome raises a question: what enables long-term phage-bacteria coexistence? We show how interactions in space can facilitate stable coexistence and long-range transport of virulent phage along with migrating bacteria. We combine large-format imaging-based experiments across multiple phage-bacteria systems with theory to reveal a chemotaxis-driven mechanism that robustly stabilizes coexistence and dispersal of virulent phages with migrating hosts. These findings suggest how spatial interaction mechanisms can stabilize antagonistic partnerships, even without cycles of defense and counter-defense, across phage-bacteria ecological systems.
Speaker: Matthias Ehrhardt
Title: Getting the most from data: Modern mathematical approaches to inverse imaging problems
Abstract: Imaging has had a tremendous impact on our society. Imaging systems such as MRI, CT or PET to diagnose and monitor disease are used everyday in our hospitals. Mathematically, getting images from such systems amounts to solving an inverse problem. Most systems use the same modelling assumptions: a static object is being probed by a single physical effect. In this talk, I want to illustrate a few approaches how we can get more out of the available data. First, what if we have data from multiple modalities available, e.g. PET and MRI? Second, what if the object of interest is moving: can we recover details in both space and time?
Speaker: Peyman Shadmani
Title: Image analysis and computer vision to inform mathematical models in cell biology
Abstract: Advances in microscopy and computational imaging are transforming how we study dynamic cellular processes. In this talk, I will present two projects that combine image analysis, computer vision, and mathematical modelling to investigate immune cell behaviour. In the first part, I will discuss a vertex-based mathematical model of phagocytosis informed by quantitative analysis of time-lapse microscopy acquired using dual micropipette experiments. Automated image analysis is used to extract physical and geometric features including engulfment cup size, pseudopod width, cell area, and phagocytosis times, enabling model calibration and refinement. In the second part, I will present a machine learning approach for identifying and quantifying microglial activation states from membrane-stained microscopy images. Using computer vision and deep learning methods, we aim to characterise changes in cell morphology associated with activation and drug response, providing quantitative tools for studying neuroinflammation and neurodegenerative disease.
Speaker: Kirsty Wan
Title: Imaging at the microscale
Abstract: Quantitative imaging and data-driven approaches are transforming our understanding of the mechanisms of cell motility and organismal behaviour. Even at the microscopic scale, we can leverage high-resolution microscopy data to capture complex and dynamic locomotion patterns and extract key physical parameters. Of particular interest are the strategies used by diverse microorganisms to control their movement, including the actuation of cilia and flagella through a fluid. These techniques can then be used to inform the development of physical models of microswimmers with varying levels of realism, providing important insights into the evolution of coordinated movement beyond nervous systems, with implications for synthetic systems and bioinspired robotics.
Speaker: Tanniemola B. Liverpool
Abstract: I will discuss some recent work looking quantitatively at the process of wound healing using ideas from thermodynamics, continuum and statistical mechanics. Wound healing is a highly conserved process required for survival of an animal after tissue damage. The wound repair process is not only of great interest in its own right but is also a laboratory to study complex tissue dynamics and regeneration.
Many wounds involve damage to an epithelial (barrier) tissue (like skin) that separates different regions of the body of a living organism. I will describe some recent work on studying wound healing in two dimensional epithelial tissues of a fruit fly pupal wing. This epithelium was chosen because it is transparent and accessible to sophisticated imaging techniques. We use live confocal time-lapse microscopy to follow the behaviour of cells in a tissue before and after wounding.
I will focus on three cell-behaviours that are generally accepted to contribute to wound re-epithelialisation: cell shape deformation, cell division, and cell migration.
I will describe how we are beginning to use a combination of mathematics, physics and biology to disentangle some of the organising principles behind the complex orchestrated dynamics that lead to wound healing.
Speaker: Joshua Bull
Title: From imaging to interactions: spatial methods for parameterising agent-based models
Abstract: Spatial proteomics and transcriptomics technologies can now capture tissue architecture in unprecedented detail, but extracting biological insight from these images demands robust quantitative methods. In this talk, we show how spatial analysis can be used to link biological imaging with agent‑based mechanistic models. We introduce MuSpAn (https://docs.muspan.co.uk/), a framework for multiscale spatial data analysis that uses mathematical and statistical tools to quantify cellular organisation and interaction. Applying MuSpAn to both images and simulations enables direct, statistically grounded comparison between model predictions and observed tissue structure, even when stochasticity affects cell positions. We demonstrate this by analysing the development of tertiary lymphoid structures in colorectal cancer, through both static imaging and dynamic modelling.