Analysis of pulse waveforms preprocessing

Pulse is one of the most important body information in Traditional Chinese Medicine (TCM). Pulse diagnosis using statistical learning theory is attracting more and more attention and taking over the traditional subjective judgments. This paper mainly presents the preprocessing algorithm of pulse waveforms. This step almost decides the whole performance of pulse analysis and pattern classification. As a result many research results about pulse preprocessing have been published, while most of the previous method only emphasize on certain aspects and hardly utilize the information in TCM pulse. We propose an integrated framework of preprocessing and introduce our method in each module. The definitions and implementations of pulse preprocessing algorithm are all presented. The proposed preprocessing method of pulse waveforms show a well performance in experiments compared with other previous methods.