Using populations of models to navigate big data in electrophysiology: Evaluation of parameter sensitivity of action potential models

• Published in: 2017 Computing in Cardiology (CinC) pp. 1 - 4
• Main Authors: Ledezma, Carlos A., Kappler, Benjamin, Meijborg, Veronique, Boukens, Bas, Stijnen, Marco, Tan, P J, Diaz-Zuccarini, Vanessa
• Format: Conference Proceeding
• Language: English
• Published: CCAL 01.09.2017
• Subjects: Computational modeling; Data models; Heart; Mathematical model; Pathology; Sociology; Statistics
• ISSN: 2325-887X
• DOI: 10.22489/CinC.2017.059-266

Experimentally-calibrated populations of models (ePoM) for cardiac electrophysiology can be used as a means to elucidate the cellular dynamics that lead to pathologies observed in organ-level measurements, while taking into account the variability inherent to living creatures. Notwithstanding, the results obtained through ePoM will depend on the capabilities of the template model, and not one model can accurately reproduce all pathologies. The objective of this work was to show how using different models, within an ePoM framework, can be advantageous when looking for the causes for a pathological behavior observed in experimental data. Populations of the ten Tusscher (2006) and the O'Hara-Rudy model were calibrated to activation-recovery intervals measured during an ex-vivo porcine heart experiment; a pathological reduction in ARI was observed as the experiment progressed in time. The ePoM approach predicted a reduction in calcium uptake via L-type channels, using the TP06 model, and an increased potassium concentration in blood, using the ORd model, as the causes for the reduction in ARI; these findings were then confirmed by other experimental data. This approach can also accommodate different biomarkers or more mathematical models to further increase its predictive capabilities.