A Novel Neural Network Model for Blood Pressure Estimation Using Photoplethesmography without Electrocardiogram

Table 4

Performance results of the linear regression, SVRM-based method, and ANN-based method with different feature extraction technique on the test database.

Method	Number of input parameters	SBP		DBP
		e (mmHg)	er (%)	e (mmHg)	er (%)
Linear regression [15]	2	9.80 ± 8.09	8.94 ± 7.57	5.88 ± 5.11	10.26 ± 8.83
RSVM based [16]	13 for SBP; 22 for DBP	5.07 ± 4.84	4.32 ± 3.59	4.31 ± 3.83	3.84 ± 3.39
ANN based (this method)	22	4.02 ± 2.79	2.84 ± 2.00	2.27 ± 1.82	4.39 ± 3.60