Establishing a Feature Pattern of Blood Stasis Syndrome with Myocardial Ischemia by Least Angle Regression

The purpose of this paper was to find some feature pattern which may be used in diagnosis of blood stasis syndrome (BSS, a unique concept of traditional Chinese medicine) with myocardial ischemia. After establishing an animal model of Chinese experimental miniature swine, 7 indicators in plasma, e.g. troponin T (cTNT), heat shock protein 27 (HSP27), cytochrome C（Cyt C）, endothelin-1 (ET-1), calcitonin gene-related peptide (CGRP), thromboxane A2 (TXA2) and prostacyclin (PGI2) were detected and then least angle regression was performed to discover the correlation between BSS and these 7 indicators. As the results showed, all the 7 indicators were contributed to the diagnosis of BSS and the regression obtained after 110 steps was curacy enough (>90%) in our cases. This pattern may become a supplementary way of traditional diagnostic methods of BBS. And it also demonstrated that least angle regression may be a better algorithm in the mining of feature patterns of syndromes.