Emergent Behavior resulting from Incorporating Cellular Uncertainty as Spatial Heterogeneity

Add to your list(s) Download to your calendar using vCal

• Richard Gray (Food and Drug Administration)
• Tuesday 04 June 2024, 14:50-15:10
• Seminar Room 1, Newton Institute.

If you have a question about this talk, please contact nobody.

FHTW02 - Fickle Heart: The intersection of UQ, AI and Digital Twins

Background: Rigorous evaluation of computational cardiac electrophysiological models is required for them be accepted for clinical use. Such models typically represent the heart as millions of ‘cardiac cells’ that exhibit the same dynamics. Rigorous model evaluation includes code verification, model validation, and uncertainty quantification (UQ) of model parameters. Recently a cardiac cellular ‘action potential (AP)’ model has been developed which includes comprehensive data-driven UQ. Unfortunately, when cellular-level UQ was propagated through the model, the resulting APs exhibited not only the desired ‘normal repolarization’ (NR) shape but also abnormal behavior such as repolarization oscillations (RO) and repolarization failure (RF). The future of robust whole heart modeling for clinical use is not ensured because of the varied behavior of cellular APs resulting from incorporating measured uncertainty into parameter values. Methods: A comprehensive model of the rabbit action potential was developed by incorporating measured cellular uncertainty as parameter distributions (as done previously for the dog). First, cellular simulations were carried out to assess the behavior of the resulting APs. Second, this cellular uncertainty was imposed as a random spatial field to investigate how such uncertainty would affect tissue-level phenomenon, including propagation, rate dependence, and spiral wave dynamics. Results: Similar to our previous results in the dog model, the APs resulting from our new rabbit model exhibited a variety of behavior with 67% exhibiting NR; 27% displaying RO; and 6% showing RF. However, when these virtual cells were paced at the rate of the typical rabbit heart, these percentages changed to: 94%: NR; 0%: RO; and 6%: RF. When this cellular uncertainty was imposed as spatial heterogeneity, the results were quite different. In 1-D and 2-D virtual tissues, wave propagation was uniform and 100% of sites exhibited normal repolarization at all pacing rates. In addition, 100 instances of reentry simulations from the same initial conditions resulted in very similar and realistic behavior:  reentrant cycle length of 120±0.5 ms  and nearly identical phase singularity trajectories, albeit with variations in time before self-termination ranging from 3 to 12 beats. Conclusion: Incorporating measured uncertainty into a model of the rabbit AP resulted in varied cellular behavior (which was rate dependent) with most cells exhibiting normal recovery at the normal heart rate. Imposing cellular uncertainty as spatial heterogeneity revealed ‘emergent robustness’ at the tissue level. Therefore, incorporating cellular uncertainty as spatial heterogeneity into whole heart models might be a feasible approach to developing rigorous models of clinically useful models of the electrical activity in the heart.

This talk is part of the Isaac Newton Institute Seminar Series series.

This talk is included in these lists:

This talk is not included in any other list

Note that ex-directory lists are not shown.