trichtu/Anatomical-awareness-in-the-electric-field-estimation

Anatomical-awareness-in-the-electric-field-estimation

There is the support file for the conference paper: anatomical awareness improves the accuracy of the real-time electric field estimation

software:

python 3.8

pytorch >=1.8

In this article, due to the trained dataset, the coil position is limited in the motor cortex region. Here is trained models for the anatomcial segmentation or electric estimation :

anatomical segmentation model:

'U_Net_Seg'

electric estimation model:

'U_Net3D','U_Net3D_Att','U_Net3D_Att_Seg','two_step'

’two_step‘ processing of electric estimation (including anatomical segmentation + electric regression model) is only processed in test mode

To achive a training process

python main.py --mode train --model U_Net3D

To achive a validation process

python main.py --mode test --model U_Net3D or python main.py --mode test --model two_step

if there is any help for your work and insights, it would be better to cite:

@INPROCEEDINGS{9533894, author={Ma, Liang and Zhong, Gangliang and Yang, Zhengyi and Fan, Linzhong and Jiang, Tianzi}, booktitle={2021 International Joint Conference on Neural Networks (IJCNN)}, title={Multi-scale anatomical awareness improves the accuracy of the real-time electric field estimation}, year={2021}, volume={}, number={}, pages={1-7}, doi={10.1109/IJCNN52387.2021.9533894}}