Research Article
ECG-Based Subject Identification Using Common Spatial Pattern and SVM
Table 3
PTB-based subject identification methods in comparison with the proposed method.
| Method | NS | DU(s) | IR | EER |
| Agrafioti and Hatzinakos [61] | 13 | 10 | 96.2 | NA | Agrafioti and Hatzinakos [62] | 14 | 5 | 100 | NA | Fatemian and Hatzinakos [63] | 13 | NA | 99.62 | NA | Safie et al. [64] | 112 | 30 | 93.60 | NA | Wang et al. [65] | 13 | NA | 84.61 | NA | Wübbeler et al. [66] | 74 | 10 | 99 | 0.03 | Zhao et al. [46] | 25 | NA | 96.00 | NA | Jekova and Bortolan [67] | 14 | 10 | 92.9 | NA | Tantawi et al. [3], Test set 1 | 14 | ~8 | 100 | NA | Test set 2 | 6 | | 83.3 | | Plataniotis et al. [41] | 14 | 10 | 100 | NA | Proposed method | 200 | 7 | | | Single-lead (I) | | | 95.15 | 0.10 | Single-lead (V3) | | | 98.92 | 0.08 | 6 limb | | | 89.89 | 0.04 | 6 chest | | | 92.08 | 0.00 | 12 leads | | | 85.54 | 0.00 |
|
|
DU: duration, NA: the information is not available or computable.
|