Skip to main content

Table 2 Performance of the proposed a wave detection algorithm on the testing dataset (APG signals measured at rest and after 2 hours of exercise)

From: Detection of a and b waves in the acceleration photoplethysmogram

  Before exercise After 2 hours of exercise
Record No of beats TP FP FN SE (%) +P(%) No of beats TP FP FN SE (%) +P(%)
A1 26 26 0 0 100 100 43 41 0 2 95.34 100
A2 24 24 0 0 100 100 47 47 0 0 100 100
B1 17 17 0 0 100 100 44 43 0 1 97.72 100
B2 26 26 0 0 100 100 38 38 0 0 100 100
C2 20 20 0 0 100 100 37 37 0 0 100 100
C3 20 20 0 0 100 100 23 23 0 0 100 100
D2 22 22 0 0 100 100 39 39 0 0 100 100
D3 19 19 0 0 100 100 27 27 0 0 100 100
E1 22 22 0 0 100 100 30 30 0 0 100 100
E2 22 22 0 0 100 100 30 30 0 0 100 100
E3 19 19 0 0 100 100 38 38 0 0 100 100
G2 30 30 0 0 100 100 49 48 0 1 97.95 100
G3 19 19 0 0 100 100 42 41 0 1 97.61 100
H3 23 23 0 0 100 100 32 32 0 0 100 100
I1 22 22 0 0 100 100 35 35 0 0 100 100
I2 17 17 0 0 100 100 31 31 0 0 100 100
J2 23 23 0 0 100 100 41 41 0 0 100 100
L2 24 24 0 0 100 100 37 37 0 0 100 100
L3 24 24 0 0 100 100 39 39 0 0 100 100
N2 18 18 0 0 100 100 24 24 0 0 100 100
N3 20 20 0 0 100 100 31 31 0 0 100 100
O1 24 24 0 0 100 100 33 33 0 0 100 100
O2 17 17 0 0 100 100 34 34 0 0 100 100
P1 26 26 0 0 100 100 34 34 0 0 100 100
P2 20 20 0 0 100 100 34 34 0 0 100 100
Q1 22 22 0 0 100 100 28 28 0 0 100 100
Q2 18 18 0 0 100 100 36 36 0 0 100 100
27 volunteers 584 584 0 0 100 100 956 951 0 5 99.57 100
  1. The PPG signals were collected from 27 subjects for 20 seconds during the 5 minutes break between each exercise [5]. To compare the performance of the proposed algorithm with the nine algorithms [5], two statistical measures were used: SE=TP/(TP+FN) and +P=TP/(TP+FP), where TP is the number of true positives (a wave detected as a wave), FN is the number of false negatives (a wave has not been detected), and FP is the number of false positives (non-a wave detected as a wave).