Because the MIMIC database contains a large variety of physiological information
data with high quality, it has been widely used in blood pressure prediction research. In
this experiment, PPG and BP signals are used, and the specific signal waveforms are
shown in Fig. 1.
3 Proposed Method
The main flow of the proposed method is as follows. Firstly, a single complete pulse
wave is extracted and normalized by an interpolation method. Then the normalized
pulse wave would be used as the input of the neural network, and finally, the estimations of blood pressure value can be obtained. Figure 2 shows the schematic
illustration of presented BP estimation frameworks.
Fig. 1. PPG and BP signal of subject 230
Fig. 2. Schematic illustration of presented BP estimation frameworks. A single pulse wave was
cut from raw pulse wave and normalized to 100 points. It is then fed into a neural network
consisting of a convolutional network and a recurrent network, and finally, the estimated blood
pressure values, namely SBP and DBP, are obtained
264
C. Wang et al.
data with high quality, it has been widely used in blood pressure prediction research. In
this experiment, PPG and BP signals are used, and the specific signal waveforms are
shown in Fig. 1.
3 Proposed Method
The main flow of the proposed method is as follows. Firstly, a single complete pulse
wave is extracted and normalized by an interpolation method. Then the normalized
pulse wave would be used as the input of the neural network, and finally, the estimations of blood pressure value can be obtained. Figure 2 shows the schematic
illustration of presented BP estimation frameworks.
Fig. 1. PPG and BP signal of subject 230
Fig. 2. Schematic illustration of presented BP estimation frameworks. A single pulse wave was
cut from raw pulse wave and normalized to 100 points. It is then fed into a neural network
consisting of a convolutional network and a recurrent network, and finally, the estimated blood
pressure values, namely SBP and DBP, are obtained
264
C. Wang et al.