Correlating serum potassium levels to ECG: A simulation study

Physiology 2012 (Edinburgh) (2012) Proc Physiol Soc 27, PC27

Poster Communications: Correlating serum potassium levels to ECG: A simulation study

S. Kharche1,2, T. Stary2, G. Callisesi2, A. Bracci2, S. Severi2

1. Department of Mathematical Sciences, University of Liverpool, Liverpool, United Kingdom. 2. DEIS, University of Bologna, Bologna, Italy.

View other abstracts by:


Alterations in serum electrolyte concentrations as caused by renal failure are associated with cardiac disease. Altered serum or extracellular potassium ([K+]o) affects cardiac electrical activity and clinical ECG. Complementary to ongoing clinical studies, this modelling study quantifies the arrythmogenic effects of abnormal [K+]o levels on electrical properties of human ventricular cells, tissue and simulated ECG. The O’Hara et al. [1] (ORD) family of epicardial (epi), midmyocardial (M) and endocardial (endo) cell models were used in this study. The ordinary differential equations were solved using in house developed robust solvers based on backward difference formulae and Newton iterations. Steady APs were elicited by pacing for 100 beats at 1 Hz after which the AP characteristics were computed for [K+]o between 2 mM and 9 mM. Dynamic AP duration (APD90) restitution was computed under basal, hypokalemic ([K+]o = 3.8 mM) and hyperkalemic ([K+]o = 5.8 mM) conditions. Regions of stable, aperiodic and alternans APs were identified. The cell models were incorporated into a 1D bidomain strand consisting of 25 endo cells, 25 M cells, and 115 epi cells [2]. The inter-cellular distance was assumed to be 0.1 mm. Implicit methods as implemented in Beatbox software were used to solve the 1D partial differential equations efficiently. A standard surface to volume ratio of 14 units was used in the simulations, while the intracellular and extracellular conductivities were optimised to give a conduction velocity (CV) of 0.44 mm/ms in a solitary wave propagating through this heterogeneous 1D strand. CV of steady waves (after 20 beats) as a function of [K+]o was computed. Further, dynamic CV restitution was computed for high and low values of [K+]o. Pseudo-ECG and APD90 dispersion were computed as a function of [K+]o. Further, the alterations of the simulated T-wave peak as well as T-wave repolarisation slope were computed and correlated to the [K+]o. The basal APD90 of endo cell model was found to be 272 ms, in M to be 332 ms, and epi to be 230 ms. APD was seen to reduce with increasing [K+]o as shown in Figure 1A. APD restitution shows the M cell type to support high amplitude alternans. An alteration of [K+]o increases the propensity of all cell types to sustain alternans at high (~ 220 to 300 ms pacing cycle length) pacing rates. In the 1D simulations, CV has a biphasic dependence on [K+]o with a maximum value at [K+]o = 4 mM. Pseudo-ECG shows a reducing T wave amplitude with increasing Ko (Fig 1B). APD dispersion also increases with an increase of [K+]o. The repolarisation slope reduced with increasing [K+]o. The effects of [K+]o on ECG were confirmed in this simulation study.



Where applicable, experiments conform with Society ethical requirements.

Site search

Filter

Content Type