Fuzzy feature for traditional Chinese medical pulse data

This paper proposed a novel method based on Wavelet Transforms which can be easily used in processing the traditional Chinese pulse data in the context of mobile healthcare. Considering the energy limitation and real-time requirements, we make a new structure to describe the pulse data which we call fuzzy feature. The fuzzy feature can extract the hiding information from the pulse units. Pulse data preprocessing and fuzzy feature extraction only operates the wavelet transform coefficients of the original data. The algorithm complexity of the fuzzy feature extraction is about O(N). Through analyzing the clusters from 28 patients’ pulse units, the fuzzy feature can extract the hiding information well. The experimental results show that the fuzzy feature can be easily used in mining useful information from patients’ data and assisting doctors to make accurate diagnosis.