The invention provides a continuous
blood pressure measurement method based on multi-parameter fusion. In addition to comprehensively considering
blood pressure calibration by using the
pulse wave transmission time, the method also considers the maximum value a01 of a first-order derivative of the blood
oxygen plethysmography which is in the period same as that of the electrocardiosignal, the
time difference value a2 of two systolic-phase peak points in two adjacent periods, the
time difference value a3 of two minimum-value points, the
time difference value a4 of two
diastole-phase peak points, the difference value a5 of two dicrotic incisure points, the amplitude a6 of the systolic-phase peak points, the amplitude a7 of the minimum-value points, the amplitude a8 of the
diastole-phase peak points, the amplitude a9 of the dicrotic incisure points, the systolic area a10, the
diastole area a11, the area a12 of the blood
oxygen plethysmography, the
area ratio a13, the difference a14 of two peak points in the same period, the difference a15 between the systolic-phase peak point and the minimum-value point in the same period, the rising time a16, the time increment a17, the
growth coefficient a18 and the
reflection coefficient a19. According to the eighteen parameters, a
blood pressure model is established by the BP neural network, the blood pressure value is predicted according to the model, a correction module is provided, the blood pressure value is accurately predicted, and the variation trend of the blood pressure is approached.