Building a Traditional Chinese Medicine-assisted Decision-making Model for Non-severe Cases of COVID-19 Based on Artificial Intelligence
Objective:To build a traditional Chinese medicine(TCM)-assisted decision-making model for non-severe cases of COVID-19 using artificial intelligence.Methods:A dataset was created by collecting medical re-cords of 314 patients with COVID-19 from the Respiratory Department,Infectious Disease Department,and Emer-gency Department of the First Affiliated Hospital of Tianjin University of Traditional Chinese Medicine from De-cember 2022 to June 2023.The dataset was split into a training set and a test set in a 7∶3 ratio.Non-severe COV-ID-19 diagnostic models were built using Random Forest,Support Vector Machine(SVM),LightGBM,and K-Near-est Neighbors,with evaluation metrics including precision,recall,Fl score,and AUC value.Results:The preci-sion of Random Forest,SVM,LightGBM,and K-Nearest Neighbors were 0.94,0.95,0.89,and 0.90,respectively.The recall values were 0.94,0.95,0.88,and 0.90,respectively,while fl scores were 0.94,0.95,0.88,and 0.90,re-spectively.The AUC values were 0.99 for all models.Conclusion:The SVM model demonstrated the highest accu-racy and is more suitable for building a TCM-assisted decision-making model for non-severe cases of COVID-19.
COVID-19Decision-making modelTraditional Chinese MedicineArtificial intelligence
万方数据
维普
