Abstract:

AbstractAn computational method based on information fusion is developed to evaluate the similarity of multiple chromatographic fingerprints of TCM,  where it solves a current  key problem on evaluating the similarity of multiple fingerprints. In this method,  the information of individual fingerprint was combined by using serial or parallel information fusion strategy at raw data level,  and then the integral similarity of multiple fingerprints was calculated by comparing their fusion results. Subsequently,  this method was applied to simulated datasets and a set of Compound Danshen Dripped Pills(CDDP) samples. The results indicate that,  by this method,  the similarity among multiple chromatographic fingerprints of TCM can be calculated,  and the lot-to-lot consistency of samples can be evaluated quantitatively.

Key words: Multiple chromatographic fingerprinting; Information fusion; Quality control of TCM