Skip to main content

Table 4 Performance evaluation with state-of-the-artwork

From: Classification of vasovagal syncope from physiological signals on tilt table testing

Authors

Results (%)

Selected features

Sample size

1. Ciliberti, M. A. P., et al. 2018 [18]

Sensitivity: 87.5

Specificity: 72.2

PPV: 75.0

NPV: 89.0

HRV, VLF, LF,

HF, LF/HF ratio

26 (all pos)

2. M. Kwok et al. 2020 [19]

Sensitivity: 82.6

Specificity: 76.8

Accuracy: 78.9

HR, CO, SV, TPR, SBP, DBP, HRV, BPV

128 (46 pos and 82 con)

3. Carmody et al. 2020 [20]

Sensitivity: 58.8

Specificity: 63.3

Accuracy: 80.2

ROC(AUC):83.2

SBP-5 s%, SBP-20 s%, HR-%I, CO- %I, SV-B2SS

101 (34 pos, 30 neg and 37 con)

4. He, Z., et al. 2021 [21]

Sensitivity: 86.0

Specificity: 82.0

ROC(AUC): 94.0

HR, RRI, SBP,

DBP, MBP,

LVET, TPR, CO, SV

203 (128 pos, 75 neg)

5. S. Hussain et al. 2022 [22]

Sensitivity: 97.6

Specificity: 92.7

Precision: 92.2

F1 score: 94.9

Accuracy: 98.9

ROC(AUC):98.3

HR, SV, CO, CI, SI, RRI, TPR, TPRI, DBP, MBP, SBP

687 (96 pos and 591 neg)

6. Proposed

Sensitivity: 87.0

Specificity: 92.7

Precision: 88.6

F1 score: 87.8

Accuracy: 90.5

ROC(AUC):95.4

CV_SBPV, CV_DBPV,

LFnu_SBPV

137 (54 pos and 83 neg)

  1. neg, test negative; con, control (healthy subject); POS, test positive; PPS, psychogenic pseudosyncope; PPV, positive predictive value; NPV, negative predictive value; RRI, R–R interval; HR, heart rate; HRV, hear rate variability; MBP, mean blood pressure; SV, stroke volume; CO, cardiac output; TPR, total peripheral resistance; LVET, left ventricular ejection time; SBPV, SBP variability; DBPV, DBP variability; SI, stroke Index; TPRI, total peripheral resistance index; (SBP-5 s%, SBP-20 s%), the percentage change in SBP from baseline at 5 s and 20 s post-stand; (HR-%I, CO- %I), the percentage increase from baseline of heart rate and CO and; (SV-B2SS), the percentage difference in SV from baseline to steady state; CV, coefficient of variance; LFnu, low-frequency normalized unit; HF, high frequency; VLF, very high frequency