基于隐结构模型和聚类分析的肺结节常见中医证候研究OA

Objective To explore the distribution of TCM syndromes of pulmonary nodules;To provide theoretical basis for standardized clinical diagnosis and treatment of pulmonary nodules.Methods This study collected general data and TCM for diagnostic information of 420 pulmonary nodules patients from Changzhou Affiliated Hospital of Nanjing University of Chinese Medicine.Factor analysis and clustering analysis were conducted using SPSS 27.0 software,while latent structure modeling was performed with the LTM-EAS algorithm in Lantern 5.0 software.Clinical summaries were combined with multivariate statistical analysis to identify common syndrome patterns of pulmonary nodules.Results A total of 175 TCM symptom items were collected.Among them,41 main symptom items with frequency>20 produced 17 common factors.The pathological locations of pulmonary nodules were primarily associated with lung,spleen and liver organs.Common pathogenic factors included deficiency,phlegm,dampness,qi,fire,stagnation and blood stasis.Combined with TCM knowledge,clustering analysis identified 8 potential TCM syndromes.Latent structure model analysis was performed on 96 symptom items with frequency≥3,and 17 latent variables were generated,yielding 9 potential syndromes through comprehensive clustering.Combined with clinical summary,7 common syndrome types and corresponding symptoms of pulmonary nodules were inferred.Conclusion Lung qi deficiency syndrome,lung spleen qi deficiency syndrome and qi yin deficiency syndrome are common in TCM deficiency syndromes of pulmonary nodules,while phlegm turbid obstructing lung syndrome,phlegm heat accumulating lung syndrome,phlegm and blood stasis internal stagnation syndrome and liver stagnation and qi stagnation syndrome are common in excess syndrome.This study uses multivariate statistical analysis method to explore the common TCM syndrome patterns of pulmonary nodules from different dimensions,and provides the basis for the standardization of TCM syndrome patterns of pulmonary nodules.