Cardiovascular Signal Processing Method Research Based-on Information Fusion

Cardiovascular disease has become the first killer of human. Use the pulse-wave signal to determine cardiac status is one of the hotspot in the biomedical engineering research. ECG, PCG and pulse-wave are important physiological signals of the cardiovascular system, they link each other closely and contain amount of physiological and pathological information. This paper not only analysis the pulse-wave but also analysis the rest two, we consider the three as a whole, research the relationship between cardiovascular status and the signals time domain and the normalized Frequency domain energy characteristics. Using algorithm and the fusion model which designed by the author integrate the signals; the data have been highly integrated. At last we give out the symbolic expression of cardiovascular status. The method of this paper can be used to prevent and control the cardiovascular disease, or evaluate the drug treatment.The main works include:(1) Analyze the signal features and Design the integration model.(2) Design information collection system of cardiovascular physiology. The system is based on virtual instrument. We designed the analog board and LabVIEW acquisition software, the hardware work well and the system easily and efficiently manages the data. Based on virtual instrument is simply and efficiently development new medical equipment;(3) Proposed a new cardiovascular physiological signal preprocessing and feature extraction method. By Kalman filtering, zero-phase filtering we improved the signal SNR; multi-sensor fusion let accurate signal segmentation and feature points location become easily. Other feature extractor methods include Lorenz plot and e energy spectrum. Finally we get the feature vectors of the cardiovascular status.(4) Through SVM we integrate the cardiovascular physiological signal and give the expression of the cardiovascular status. Experiments demonstrate Authenticated the model.