Recognition of sub-health state based on new pulse and ECG information

In order to evaluate sub-health state, synchronous acquisition of pulse images and ECG signals multi-information was conducted by using self-designed synchronous acquisition system. Stable pulse waves were obtained through the grid area change of each frame pulse images. After the three dimensional pulse image was reconstructed for every frame image, pulse length and pulse width were calculated. The pulse and ECG features such as power spectrum peak, gravity frequency, and approximate entropy were extracted. Linear Discriminant Analysis (LDA) algorithm was used for sub-health state classification. The satisfactory recognition result was obtained by using all ECG and pulse features, which provide a new research approach for evaluating sub-health state.