Analysis of syndrome discipline of generalized anxiety disorder using data mining techniques

OBJECTIVE: To study the use of data mining techniques in analyzing the syndrome discipline of generalized anxiety disorder (GAD). METHODS: From August 1, 2009 to July 31, 2010, 705 patients with GAD in 10 hospitals of Beijing were investigated over one year. Data mining techniques, such as Bayes net and cluster analysis, were used to analyze the syndrome discipline of GAD.RESULTS: A total of 61 symptoms of GAD were screened out. By using Bayes net, nine syndromes of GAD were abstracted based on the symptoms. Eight syndromes were abstracted by cluster analysis. After screening for duplicate syndromes and combining the experts’ experience and traditional Chinese medicine theory, six syndromes of GAD were defined. These included depressed liver qi transforming into fire, phlegm-heat harassing the heart, liver depression and spleen deficiency, heart-kidney non-interaction, dual deficiency of the heart and spleen, and kidney deficiency and liver yang hyperactivity. Based on the results, the draft of Syndrome Diagnostic Criteria for Generalized Anxiety Disorder was developed. CONCLUSION: Data mining techniques such as Bayes net and cluster analysis have certain future potential for establishing syndrome models and analyzing syndrome discipline, thus they are suitable for the research of syndrome differentiation.