Table 2. Comparisons between SampEn and other typical nonstationary and nonlinear analysis methods.
It can be seen that the RQA analysis generally presents negative correlation with BIS although DET has relatively higher correlation with BIS. Compared to entropy methods, the AUC and MAE show inferior values. Both SampEn and PEn have high AUC values, however, the SampEn shows better correlation.
| CC | AUC | MAE | |
|---|---|---|---|
| RR (RQA) | −0.16 ± 0.16 | 0.24 ± 0.13 | 48.60 ± 7.95 |
| DET (RQA) | −0.39 ± 0.16 | 0.16 ± 0.10 | 58.33 ± 7.98 |
| PEn | 0.55 ± 0.24 | 0.91 ± 0.09 | 42.81 ± 7.92 |
| SampEn | 0.72 ± 0.09 | 0.95 ± 0.04 | 47.23 ± 7.84 |
Notes.
MAE is not an absolute metric because it can decrease through regression. Values are presented with mean ± STD. Markers symbolize the statistical significant Student t-test.
- PEn
- Permutation
- CC
- correlation coefficient
- AUC
- area under curve
- MAE
- mean absolute error
- RQA
- recurrence quantification analysis
- RR
- recurrence rate
- DET
- determinism
- SampEn
- sample entropy