Research Article
Real-Time Arrhythmia Classification Algorithm Using Time-Domain ECG Feature Based on FFNN and CNN
Table 2
The description of the constructed two datasets.
| Dataset | ECG_200 ms_POS_72 | ECG_RRR_TYPE_360 |
| Input | 200 ms ECG segment (mV) | “ECG_RRR” feature (mV) | Input shape | 72 | 360 | Output | QRS complex position ([0, 71]) | Arrhythmia type ([0, 12]) | Output shape | 1 | 1 | Total samples | 794376 | 97898 |
|
|