Presentation Information
[C-15-15]Group-Level Evaluation of Staircasing Error Reduction Using a Tensor-Based SPFD Method with Individual Brain Models
◎△Eikei Yamada1, Akimasa Hirata1 (1. Nagoya Institute of Technology)
Keywords:
Scalar-Potential Finite-Difference Method,Computational accuracy,Staircasing error
In low-frequency magnetic field dosimetry using voxel-based human models, staircasing errors arising from the stepwise approximation of curved tissue boundaries remain a key issue. We have proposed a tensor-based three-dimensional SPFD method that accounts for the boundary conditions at tissue interfaces, and demonstrated its effectiveness using multilayer spheres and standard head models. In this study, the method was applied to 30 individual brain models, and the reduction of the induced electric field on the gray-matter gyral surface was evaluated at the group level under uniform and TMS exposure. A consistent reduction relative to the conventional method was obtained in all 30 models (about 3–8 % for the 99th-percentile and about 11–15 % for the maximum value), and was most pronounced in the high-field tail. These results indicate that the staircasing-error reduction by the proposed method is reproducible across individual brain models.