Issue with simulation of Temporal Interferece on a head model

[Paria M.]

Hi,

I'm running a simulation of Temporal interference on a four-layer head model. There are two pairs of electrodes attached to the head model, one with a 1 KHz electrical wave and the other with a 1.04 KHz frequency. However, I'm encountering an issue with the placement of the electrodes on the scalp.

When observing the overall field, I noticed two regions within the simulation where there is no field, indicating a problem with the electrode placement. I've attached a picture of the observed field for reference.

Here's a summary of my project progress so far:

1. Created four spheres representing the layers of the head model: brain, CSF, skull, and scalp.
2. Subtracted the inner spheres from the outer ones to accurately represent the head's structure.
3. Positioned four cylinders along the radial axis of the spheres to represent the electrode pairs. The cylinders are aligned with the radius of the spheres and intersect with two outer layers.
4. Attempted to attach the electrodes to the scalp by subtracting the intersected layers from the cylinders. However, the attachment is not as expected.
5. Started a new quasi-static EM LF simulation
6. Configured the simulation setup, materials (for brain layers), boundary conditions (for electrodes), grade, and voxel settings for each pair of electrodes separately.
7. Assigned a priority of 1 to brain layers and a priority of 2 to electrodes to prioritize brain tissues during simulation voxelization.
8. Extracted the overall field for both electrode pairs and selected them to observe the maximum modulation field.

I would appreciate any insights or ideas on why this issue is occurring.

Thank you!

picture of the observed field

[AntoninoMC]

Dear @Paria-M,
thank you for the detailed description of the issue. I think that there is an issue in the voxeling, as the electrode seem to have been voxeled within the head layers. We rarely have seen such an issue, and we will investigate more. However, for the moment, I can suggest you a more robust way to proceed:

1. When it is not necessary, avoid executing boolean operations: for example, you can create you multi-shell head model creating multiple sphere of different diameter, and you can play with voxel priorities. In this project, I would have created the four spheres for the brain, CSF, skull and scalp, and then I would have created five voxel priorities for electrodes (priority 0), scalp (priority 1), skull (priority 2), CSF (priority 3) and brain (priority 4). The larger priorities dominate over the smaller ones, therefore the electrode geometries would have no penetrated the skin, etc.
2. The same procedure in 1. applies for the electrodes. You can create cylinders that penetrate the head, and then put the priority of these electrodes smaller than the priorities of all the other head tissues.

This approach is efficient also because permits you to simply modify the model geometry if necessary (e.g. to change the thickness of one shell, or electrode geometry) without the need to built again the model from scratch. In addition, if in the future you would like to create unstructured meshes (alternative discretization method provided by Sim4Life) on the model geometry, you should not find difficulties (boolean operation may introduce defects in the model geometry).

I hope that these suggestions will be useful!

[Paria M.]

Dear AntoninoMC,

Thank you for your comprehensive solution!!!
Following your advice, I assigned distinct priorities to each tissue, and there is no more need for boolean operations (subtraction). This approach proved to be completely successful.