In this pull request there are a few changes and additions to the solver. They are summarized below:
1) ToRORd_Land model:
A new variation of the ToRORd human cellular model to the repository has been implemented. It is called ToRORd_Land and it is a mix between the original ToRORd with the Land model. The model itself is suitable for electromechanical simulations, but we can control the fibre stretch of the Land model by setting the 'lambda' parameter equal to 1 (no deformation) leading to a 0D model.
In all my tests (single cell, cable, plain and biventricular) the model can be solved using a fixed Rush-Larsen timestep of dt=0.01ms without NaN errors and with an accuracy closer to the original Matlab code. Also, the user can include current modifiers to the model by using the new extra data functions "_set_extra_data_mixed_torord_Land_samecelltype" and "_set_extra_data_mixed_torord_Land_epi_midendo" and change the AP behaviour as desired. To test the model capabilities three new configuration files have been added to the _exampleconfigs/ folder:
"cable_mesh_with_ToRORd_Land_endo_mid_epi.ini"
"plain_mesh_ToRORd_Land_endo_mid_epi.ini"
"trace_state_vector_cell_ToRORd_Land.ini"
2) Change the PMJ delay calculus:
The PMJ delay calculus is now made using the difference between the LAT of the terminal Purkinje volume and the minimum LAT obtained from the coupled ventricular volumes to that Purkinje terminal. We no longer use a mean LAT of the coupled ventricular volumes.
3) New script to read ".acm" files:
Reader script to extract the data from a ".acm" file has been added to the /scripts folder.
Hello Sachetto,
In this pull request there are a few changes and additions to the solver. They are summarized below:
1) ToRORd_Land model:
A new variation of the ToRORd human cellular model to the repository has been implemented. It is called ToRORd_Land and it is a mix between the original ToRORd with the Land model. The model itself is suitable for electromechanical simulations, but we can control the fibre stretch of the Land model by setting the 'lambda' parameter equal to 1 (no deformation) leading to a 0D model.
In all my tests (single cell, cable, plain and biventricular) the model can be solved using a fixed Rush-Larsen timestep of dt=0.01ms without NaN errors and with an accuracy closer to the original Matlab code. Also, the user can include current modifiers to the model by using the new extra data functions "_set_extra_data_mixed_torord_Land_samecelltype" and "_set_extra_data_mixed_torord_Land_epi_midendo" and change the AP behaviour as desired. To test the model capabilities three new configuration files have been added to the _exampleconfigs/ folder:
2) Change the PMJ delay calculus:
The PMJ delay calculus is now made using the difference between the LAT of the terminal Purkinje volume and the minimum LAT obtained from the coupled ventricular volumes to that Purkinje terminal. We no longer use a mean LAT of the coupled ventricular volumes.
3) New script to read ".acm" files:
Reader script to extract the data from a ".acm" file has been added to the /scripts folder.