Adding electrodes to the Grid changes simulation results

[tbone]

Hello,

I was following the Tutorial:
F.Temporal Interference Stimulation with Ohmic Electro Quasi-Static Solver with a Personalized Head Model

Here I've noticed for the 'Grid' settings:

• The bounding box (or wire block?) for the head was used for manual gridding with 0.7 mm resolution
• Automatic Grid settings were set for the 4 TI electrodes

I had some questions in regard to this setup.

1. When I do not add the Automatic Grid settings for the 4 TI electrodes (so just one manual grid for the bounding box), the interfering E-field results change. I am confused why there is a change, as the bounding box still encompasses the head model and the electrodes.
2. Why not just include the entire head model instead of the bounding box? Would it not be more accurate to automatically grid the segmented surfaces of the head model?

Comparing the results for the different grid settings was done in the same manner by:
Overall Field - Normalize Frequency Domain Results (0.001 A) - Multiplier (EM x2) - Mask Filter to ROI - Max Modulation

Thank you for your time and help!