Table 7.
Statistical results for comparison between normal and abnormal heart sound recordings on all selected balanced data from training set.
| Feature | Abnormal | Normal | P-value |
|---|---|---|---|
| m_RR (ms) | 919 ± 279 | 876 ± 232 | <0.05 |
| sd_RR (ms) | 54 ± 40 | 44 ± 30 | <0.01 |
| m_IntS1 (ms) | 131 ± 9 | 130 ± 8 | <0.01 |
| sd_IntS1 (ms) | 18 ± 6 | 15 ± 5 | <0.01 |
| m_IntS2 (ms) | 105 ± 10 | 106 ± 10 | 0.6 |
| sd_IntS2 (ms) | 16 ± 6 | 13 ± 5 | <0.01 |
| m_IntSys (ms) | 213 ± 112 | 199 ± 82 | <0.05 |
| sd_IntSys (ms) | 20 ± 9 | 15 ± 7 | <0.01 |
| m_IntDia (ms) | 470 ± 216 | 442 ± 176 | <0.05 |
| sd_IntDia (ms) | 39 ± 29 | 33 ± 22 | <0.01 |
| m_Ratio_SysRR (%) | 23 ± 3 | 23 ± 3 | 0.9 |
| sd_Ratio_SysRR (%) | 4.0 ± 1.7 | 3.5 ± 1.6 | <0.01 |
| m_Ratio_DiaRR (%) | 44 ± 5 | 44 ± 5 | 0.3 |
| sd_Ratio_DiaRR (%) | 6.4 ± 2.9 | 5.7 ± 2.6 | <0.01 |
| m_Ratio_SysDia (%) | 55 ± 12 | 54 ± 12 | 0.2 |
| sd_Ratio_SysDia (%) | 15 ± 6 | 13 ± 6 | <0.01 |
| m_Amp_SysS1 (%) | 48 ± 22 | 44 ± 22 | <0.01 |
| sd_Amp_SysS1 (%) | 29 ± 16 | 24 ± 19 | <0.01 |
| m_Amp_DiaS2 (%) | 64± 24 | 60 ± 27 | <0.01 |
| sd_Amp_DiaS2 (%) | 37 ± 23 | 33 ± 19 | <0.01 |