| 21 |
How active are your heart cells?Heart state estimation from electrocardiograms 
|
|---|
This study addresses the estimation of parameters affecting the shape of ECG waveforms. The parameters include electrical conduction velocities and potential activation patterns that control the behavior of cardiomyocytes. Although the broad correspondence between diseases and ECG waveforms has been well studied, it has been difficult to estimate cardiac status at the cellular level from an observed ECG. This study enables its estimation using an artificial neural network trained with a set of ECGs generated by an existing ECG simulator. It is expected that the estimated parameters can be used for detailed and individualized cardiac status analysis. After evaluating this method in a clinical environment, we want to introduce this technique to our bio-digital twin that can predict personalized health status and simulate medical care.
[1] S. Sugiura, T. Washio, A. Hatano, J. Okada, H. Watanabe, T. Hisada, “Multi-scale simulations of cardiac electrophysiology and mechanics using the university of Tokyo heart simulator,” Progress in Biophysics and Molecular Biology, Vol. 110, pp. 380–389, 2012.
[2] R. Nishikimi, M. Nakano, K. Kashino, S. Tsukada, “Variational autoencoder–based neural electrocardiogram synthesis trained by FEM-based heart simulator,” Cardiovascular Digital Health Journal, Vol. 5, Issue 1, pp. 19-28, 2024.
Ryo Nishikimi
Biomedical Informatics Research Group, Media Information Laboratory