CHIMERA visit to the University of Exeter

8 July 2024

On June 25, 2024, a team of early career researchers from CHIMERA attended the Datathon Conference: “Quantifying heterogeneity and variability in human daily rhythms”, organised by the EPSRC Hub for Quantitative Modelling in Healthcare (HQMH) at the University of Exeter.

The CHIMERA team pictured in front of the Living Systems Institute

The organisers provided time-series data of multiple hormones from 214 individuals, sampled every 20 minutes over 24 hours. Data on sex, age, BMI, blood pressure and behavioural markers were also made available. The aim of the exercise was to quantify the contribution of physiological and behavioural parameters to hormonal variability. The competition was divided into three categories: Best Predictive Inference, Best Novelty of Approach and Best Negative Result. The CHIMERA team, formed of Vijay Anand, James Butterworth, Jun Cheng, Gevik Grigorian and Manuel Teixeira won first place for Novelty of Approach. CHIMERA academic Alejandro Diaz was part of the discussions for the teams submission and was present during the final presentations in Exeter. Congratulations to all the CHIMERA team members involved!

The datathon was followed the next day by a joint University of Exeter-UCL Early Career Researcher (ECR) event that included a showcase of mathematical research from both UCL CHIMERA and the Exeter HQMH. Nick Ovenden introduced the CHIMERA hub and our seven ECR speakers: Vijai Anand, Manuel Teixeira Cabeleira, Jun Cheng, James Butterworth, Gevik Grigorian, Abi Smith and Finn Catling. Vijai spoke first about his research utilising a mechanical ventilation model to assess the effectiveness of airway clearance techniques. He was followed by Manuel, who presented a fully coupled cardiopulmonary model incorporating a novel alveoli opening mechanism. Next was two members of the data science team, starting with Jun Cheng discussing how dynamic Bayesian history matching can be used to calibrate a patient-specific lung model; Jun’s talk was followed by James Butterworth,  explaining how so-called indirect encodings can be used to morph the search manifolds of the biomechanical models. Gevik spoke after James on his research in advanced machine learning techniques, detailing how scientific machine learning can learn missing components of biomechanical models from given data. CHIMERA’s final two talks then presented specialised computational approaches for specific clinical applications. Abi Smith presented a novel poroelastic lung model that can be used to better understand gravitational effects on patient positioning in intensive care. Finally, Finn discussed how analysis of an arterial waveform model could enable more personalised sepsis treatment. Our showcase was closed by a brief presentation from Alex Diaz, who talked about potential future research directions beyond CHIMERA.

For the remainder of the ECR day, we learnt about the exciting research being done in the Exeter HQMH and listened to some excellent talks on career development both inside and outside academia. At the end of the day, Alex Diaz and Nick Ovenden sat on a Q&A panel answering varied questions from the ECR audience. The HQMH team at University of Exeter were fantastic hosts and we are most grateful for this valuable opportunity to host a joint event with them.