Fig 1.
Schematic representation of our thin 3D tissue slab model along with the 10 design input parameters considered: Scar radius (SR), Scar depth (SD), Scar conductivity (SC), Engineered heart tissue thickness (EHTt), Engineered heart tissue conductivity (EHTc), Contact area (CA), Conductive polymer thickness (CPt), Conductive polymer conductivity (CPc), Internal bath thickness (IBT), Tissue thickness variation (Δtt).
Table 1.
Parameter ranges from literature for the 10 model input parameters.
Fig 2.
Value ranges of the 10 parameters (rows) used in the two versions of our model (Epi-endo, and Endo-epi, 1st and 2nd columns respectively). The 3 sub-cases (Transmural, Block and Fixed) are shown in the 3rd, 4th and 5th columns. The table shows which parameters of the original model are fixed in order to derive each sub-case. The bottom row displays representations of each model version and subcases.
Table 2.
Model quantities kept fixed in all simulations: conductivity of the internal and external bath areas, ionic model used for the EHT, conductivities and fiber orientations for the host myocardium.
Table 3.
Modifications to model parameters to match the experimental studies setups.
The first row reports the ranges used in our simulation study, while the rows below display the parameters values used to replicate each of the 3 experiments.
Fig 3.
Results of the model validation.
Experiments from 4 papers were compared to the model: Jackman et al (A), Thompson et al (B), Zimmermann et al (C) and Mawad et al (D-E). In panel A CV is evaluated in 3 different areas: on the patch, on the tissue far from the patch and on the tissue under the patch. In all 3 experiments, CV before (control) and after (graft) the attachment of the patch is compared. The CV measured experimentally (mean and standard deviation) are represented as black dots with error bars, with the p-values associated with each experiment. The simulated CV are displayed as red diamonds.
Fig 4.
Cross-species comparison of the 500 simulations run to create the training datasets, classified as lower half, upper half or EHT, according to the propagation path (arrows in the left panel) followed by the stimulus.
The colours on the right panel represent the arrows colours in the left one. Yellow, green and blue columns show the percentage of simulations classified as lower half, upper half or EHT, respectively. The percentage values are reported on each column.
Fig 5.
Donut charts representing the total effect index from the GSA.
The figure compares the total effect indices obtained from the rat, rabbit, and human models (rows), for the epi-endo and endo-epi versions of the model and the transmural, block and fixed model sub-cases (see section 2.1).
Fig 6.
Upper row: propagation in a healthy slab. Lower row: propagation in a sample simulation of the fixed setup, showing the delay in REAT activation.
Fig 7.
Model output (REAT) VS EHT conductivity for the rat, rabbit and human model, in the fixed setup.
Boxplots represent the distribution of REATs among the 500 LH simulations that had the EHT conductivity restricted to current experimental ranges. Diamonds represent REATs obtained by fixing all the other parameters and increasing EHT conductivity by multiplying the upper bound of the range used in the LH by 1.5, 2, 2.5, 3, 3.5, 4 and 4.5 folds. Dashed line indicates the healthy REAT estimated by our models.
Fig 8.
Summary of the 4 cases tested for arrhythmic behavior.
For the Baseline and High EHT cases, EHTc is set to 0.112 S/m (mid value in experimental range) and 1.008 S/m, respectively, and no modifications to ion channels’ densities are made. When the density of the ion channels responsible for the fast Sodium current and the inward rectifier potassium current are increased, the EHTc is set to 0.112 S/m.The left panel shows which cases match estimated healthy REATs. The center panel shows the distribution of the mean repolarization gradients at the interface between myocardium and EHT (on 100 beats pacing protocol). The right panel shows the presence or absence of ectopic beats fired from the EHT.