Skip to main content
. 2013 Nov 19;15(1):104. doi: 10.1186/1532-429X-15-104

Table 3.

ICA-based R-peak detection results for the training and test datasets using different lead configurations

 
Configuration
Training dataset
Test dataset
LC # leads # PCs Leads I, II + Se [%] + P [%] μ pd [ms] σ pd [ms] Se [%] + P [%] μ pd [ms] σ pd [ms]
1
8
8
V1-V6
99.5
99.4
3.5
6.6
99.2
99.1
5.8
5.0
2a
7
6
V2-V6
99.2
99.1
5.9
9.6
98.9
99
12.1
8.2
2b
6
6
V1-V4
99.2
99.3
4.5
5.6
99.4
98.3
11.7
13.3
2c
5
4
V2-V4
99.2
99.2
4.5
10.3
98.7
97
14.5
13.6
2d
4
4
V3-V4
98.9
98.5
4.5
6.9
98.3
97.7
16.9
13.4
2e
3
3
V2
94.7
94.5
12.2
10.5
95.2
93.7
12.1
15.9
2f
3
3
V3
97.3
92.8
11.7
8.5
96.5
92.4
14.5
9.7
2g
3
3
V4
98.2
97.3
6.3
11.5
98.5
97.2
13.9
12.2
2h 2 2 77.1 66.3 33.3 20.6 78.9 65.8 22.1 11.2

R-peak detection results for the training and test datasets using the ICA-based method with lead configurations LC2a-LC2g. Averaged results of lead configuration LC1 are shown for comparison (μpd: mean detection delay; σpd: jitter).