Table 6.
Comparison results with prior work. BCG: ballistocardiography, ECG: electrocardiogram, IPG: impedance plethysmography, PPG: non-invasive photoplethysmography.
| Calibration | Subjects | Acquired Signals (Measure Location) |
Accuracy w.r.t. cuff BP [mean ± SD (MAD)/mmHg; r] |
|
|---|---|---|---|---|
| Method | Intervention | |||
| Probability distributions [31] |
Daily activities | N = 51 | Only ECG: chest | SBP:/± 10.22 (7.72);/ DBP:/± 10.03 (9.45);/ |
| oPTP [32] | Supine position | N = 26 | Only PPG: ear, toe | SBP: 2.16 ± 6.23 (8†);/ DBP: −1.49 ± 6.51 (9†);/ |
| oPTP [33] | Deep breathing + Sustained handgrip |
N = 15 | ECG: chest BCG: foots |
SBP: −0.03 ± 8.58 (9) †; 0.70 DBP: −0.02 ± 5.81 (7) †; 0.66 |
| oPTP [34] | handgrip exercises | N = 15 | IPG: wrist PPG: finger |
SBP: 0.31 ± 8.55 (5†); 0.88 DBP: −0.5 ± 5.07 (4†); 0.88 |
| PTP, three pairs [35] | Post-exercise | N = 10 | PPG: finger Bio-Z: wrist |
SBP: −0.09 ± 7.16 (6†); 0.86 DBP: −0.02 ± 2.56 (5†); 0.77 |
| Multi-wavelength [36] | Deep breathing + Valsalva maneuver |
N = 40 | ECG: wrist, foot PPG: finger ICG: neck, thorax |
SBP: 0.00† ± 2.85 (2.2); 0.98 DBP: 0.00† ± 1.75 (1.4); 0.99 |
| Machine learning [22] | MIMIC-III [37] | N = 155† | ECG and PPG [37] | SBP: 0.03 ± 5.52 (3.27); 0.97 DBP: 0.10 ± 1.97 (1.16); 0.97 |
| mPTP, this work (dMK-BH model) |
Running exercise | N = 21 | ECG: wrist, foot PPG: finger |
SBP: −0.77 ± 7.79 (6.0); 0.83 DBP: 0.51 ± 5.70 (4.4); 0.74 |
| fPTP, this work (dMK-BH model) |
Running exercise | N = 21 | ECG: wrist, foot PPG: finger |
SBP: −0.54 ± 6.95 (5.3); 0.87 DBP: 0.24 ± 5.21 (4.0); 0.78 |
Note 1: “/” Not be estimated based on reported results or able to be reported in corresponding authors’ other work. Note 2: “†” Be approximately estimated from the corresponding Bland–Altman plots.