Table 6 Recognition rate and kappa coefficient in other research
From: A transition-constrained discrete hidden Markov model for automatic sleep staging
(a) GOHMM[5](kappa = 0.36) | ||||||
|---|---|---|---|---|---|---|
Manual scoring | Â | Wake | S1 | S2 | SWS | REM |
| Â | Wake (%) | 86 | 11 | 0 | 0 | 3 |
| Â | S1 (%) | 52 | 22 | 6 | 7 | 13 |
| Â | S2 (%) | 13 | 12 | 14 | 24 | 37 |
| Â | SWS (%) | 1 | 0 | 4 | 95 | 0 |
| Â | REM (%) | 32 | 16 | 13 | 13 | 26 |
(b) GOHMM[6]( kappa = 0.50) | ||||||
Manual scoring | Wake (%) | 79 | 10 | 4 | 0 | 7 |
| Â | S1 (%) | 21 | 24 | 19 | 5 | 31 |
| Â | S2 (%) | 3 | 8 | 36 | 24 | 29 |
| Â | SWS (%) | 0 | 0 | 6.5 | 93.5 | 0 |
| Â | REM (%) | 14 | 13 | 4 | 1 | 68 |
(c) HMM[7]( kappa = 0.52) | ||||||
Manual scoring | Wake (%) | 51.04 | 47.30 | 0.41 | 0 | 1.24 |
| Â | S1 (%) | 0 | 4.84 | 48.39 | 0 | 46.77 |
| Â | S2 (%) | 0 | 0 | 68.62 | 29.34 | 2.04 |
| Â | SWS (%) | 0 | 1.09 | 0 | 98.91 | 0 |
| Â | REM (%) | 0 | 0.94 | 4.72 | 8.02 | 86.32 |