Research Article
[Retracted] An ECG Heartbeat Classification Method Based on Deep Convolutional Neural Network
Table 6
Comparison of VEB and SVEB classification performance of the proposed method with other methods.
| | Methods | Classifier | VEB | SVEB | | Acc | Sen | Spe | Ppr | Acc | Sen | Spe | Ppr |
| | Jiang et al. [26] | BBNN | 98.8 | 94.3 | 99.4 | 95.8 | 97.5 | 74.9 | 98.8 | 78.8 | | Kiranyaz et al. [27] | 1D-CNN | 99.0 | 93.9 | 98.9 | 90.6 | 97.6 | 60.3 | 99.2 | 63.5 | | Acharya et al. [28] | DA + CNN | 97.9 | 94.2 | 98.8 | 95.3 | 97.0 | 90.6 | 98.6 | 94.3 | | Zhai et al. [29] | 2D-CNN | 99.1 | 96.4 | 99.5 | 96.4 | 97.3 | 85.3 | 98.0 | 71.8 | | Shaker et al. [30] | GAN + CNN | 99.5 | 94.5 | 99.7 | 98.6 | 99.1 | 91.2 | 99.3 | 97.7 | | Proposed | 1D-CNN | 99.8 | 98.4 | 99.9 | 98.5 | 99.7 | 92.1 | 99.9 | 96.8 |
|
|
BBNN: block-based neural networks, 1D: one-dimensional, 2D: two-dimensional, CNN: convolutional neural networks, DA: data augmentation, and GAN: generative adversarial networks.
|
