Table 2 Field error values from Eq. (6) for the mid-surface between gray and white matter.
From: A fast direct solver for surface-based whole-head modeling of transcranial magnetic stimulation
Error type and percentage values | SimNIBS Ernie model (%) | 4 Connectome models, averaged headreco seg | 4 Headreco models, averaged FreeSurfer seg |
|---|---|---|---|
FMM LU precision: 1e−4: Field FMM precision: 1e−4 | |||
\({Error}_{total}\) for entire midsurface/BEM-FMM LU versus iterative sol | 0.10 | 0.11%0.02% | 0.12%0.02% |
\({Error}_{total}\) for 4 cm ROI under coil/BEM-FMM LU versus iterative sol | 0.03 | 0.03%0.03% | 0.05%0.02% |
\({Error}_{mag}\) for entire midsurface/BEM-FMM LU versus iterative sol | 0.08 | 0.08%0.01% | 0.09%0.01% |
\({Error}_{mag}\) for 4 cm ROI under coil/BEM-FMM LU versus iterative sol | 0.03 | 0.02%0.005% | 0.05%0.02% |
FMM LU precision: 1e−4: Field FMM precision: 1e−1 | |||
\({Error}_{total}\) for entire midsurface/BEM-FMM LU versus iterative sol | 0.37 | 0.38%0.02% | 1.34%0.29% |
\({Error}_{total}\) for 4 cm ROI under coil/BEM-FMM LU versus iterative sol | 0.15 | 0.16%0.02% | 0.40%0.08% |
\({Error}_{mag}\) for entire midsurface/BEM-FMM LU versus iterative sol | 0.28 | 0.27%0.02% | 1.00%0.15% |
\({Error}_{mag}\) for 4 cm ROI under coil/BEM-FMM LU versus iterative sol | 0.15 | 0.14%0.01% | 0.28%0.03% |
FMM LU precision: 1e−3: Field FMM precision: 1e−1 | |||
\({Error}_{total}\) for entire midsurface/BEM-FMM LU versus iterative sol | 1.35 | 1.46%0.29% | 0.90%0.98% |
\({Error}_{total}\) for 4 cm ROI under coil/BEM-FMM LU versus iterative sol | 0.40 | 0.42%0.16% | 0.20%0.21% |
\({Error}_{mag}\) for entire midsurface/BEM-FMM LU versus iterative sol | 1.05 | 1.04%0.19% | 0.61%0.69% |
\({Error}_{mag}\) for 4 cm ROI under coil/BEM-FMM LU versus iterative sol | 0.39 | 0.37%0.18% | 0.15%0.16% |
