Table 1 Average removal performance of different artifacts at different signal-to-noise ratio levels.
From: A novel EEG artifact removal algorithm based on an advanced attention mechanism
Artifact type | Method | Loss | SNR | RRMSEt | RRMSEf | CC |
---|---|---|---|---|---|---|
EMG | 1D-ResCNN | 0.096 | 7.474 | 0.393 | 0.444 | 0.891 |
NovelCNN | 0.069 | 11.570 | 0.342 | 0.319 | 0.934 | |
DuoCL | 0.061 | 11.670 | 0.286 | 0.308 | 0.933 | |
CLEnet | 0.059 | 11.929 | 0.283 | 0.300 | 0.935 | |
EOG | 1D-ResCNN | 0.062 | 11.410 | 0.311 | 0.309 | 0.936 |
NovelCNN | 0.084 | 11.344 | 0.356 | 0.338 | 0.921 | |
DuoCL | 0.059 | 12.881 | 0.270 | 0.284 | 0.939 | |
CLEnet | 0.060 | 12.337 | 0.274 | 0.287 | 0.934 | |
EMG + EOG | 1D-ResCNN | 0.098 | 7.383 | 0.403 | 0.447 | 0.889 |
NovelCNN | 0.084 | 10.950 | 0.372 | 0.346 | 0.917 | |
DuoCL | 0.070 | 11.388 | 0.307 | 0.320 | 0.925 | |
CLEnet | 0.068 | 11.498 | 0.300 | 0.319 | 0.925 |