(a) RDM 42 points per layer

(b) MAG with 42 points per layer

(c) RDM with 162 points per layer

(d) MAG with 162 points per layer

(e) RDM with 642 points per layer

(f) MAG with 642 points per layer

Figure 11: Forward EEG: accuracy comparison of different BEM solvers with three-layer sphere head models. We observe that the symmetric BEM outperforms the other BEM methods in term of precision. The TFEM was run on a mesh with 427,000 vertices, and the results were interpolated to 162 points in (c) and (d) and 642 points in (e) and (f).