Research Article
Arrhythmia Classification of ECG Signals Using Hybrid Features
Table 4
Comparison of the proposed scheme with state-of-the-art methods using class-oriented scheme.
| | Features | Dimension | Classes | Accuracy | Sensitivity | Specificity | PPV |
| Proposed | DWT + RR + TEO | 17 | 18 | 99.75 | 98.7 | 99.9 | 99.8 | Zidelmal et al. [8] | Frequency content + RR + QRS | 13 | 2 | 97.2 | 99 | — | — | Ye et al. [11] | WT + ICA + RR | 18 | 16 | 99.3 | 91.3 | — | — | Ebrahimzadeh et al. [4] | HOS + timing interval | 24 | 5 | 95.18 | 95.61 | 98.8 | 90.6 | Pathoumvanh et al. [24] | DCT | 5 | 5 | 99.11 | 97.01 | 99.44 | — | Rabee and Barhumi [26] | Multi resolution WT | 251 | 14 | 99.2 | 96.2 | 100 | — | Alajlan et al. [27] | HOS of 2nd-order-cumulant | 604 | 2 | 94.96 | 92.19 | 95.19 | — | de Oliveira et al. [14] | Waveform + RR | — | 2 | 95 | 95 | 99.87 | 98 | Li et al. [19] | Timing interval + waveform amplitude | — | 2 | 98.2 | 93.1 | — | 81.4 |
|
|