Harry Saxton
I am a final year PhD student from Sheffield Hallam University
Research Summary: Lumped parameter models offer an efficient alternative to assessing a patient’s physiological state, thus
making their utilisation in digital twins appealing. Before the models are utilised on clinically relevant data, one has to understand the dynamics, uncertainty and identifiability embedded in the model, leveraging methodologies in global sensitivity
analysis, orthogonality analysis, profile likelihood and Kalman filtration as means of assessment. The theoretical underpin‑
ning conducted in this work ensures unique, identifiable patient‑specific input parameters.
Session
Personalised computational models of cardiovascular physiology have the potential to revolutionise patient-specific diagnosis and prediction. This study introduces a novel workflow that efficiently identifies the optimal bio-markers for cardiovascular personalisation, leveraging ModellingToolKit.jl, GlobalSensitivity.jl and QuasiMonteCarlo.jl. The outcome of our research is a more reliable and more practically focused personalisation of complex cardiovascular models.