Abstract: Objective To investigate the diagnostic value of thromboelastogram parameters (TEG) combined with clinical indexes for acute ischemic stroke (AIS). Methods A total of 190 patients were enrolled in this study, including 120 acute ischemic stroke patients in the observation group and 70 other patients as the controls. Complete blood count, coagulation function, D-dimer, fasting blood glucose (GLU) and blood lipid parameters were analyzed. Lasso regression and logistic regression were used to screen variables for AIS diagnostic model presented by nomogram. Receiver operating characteristic (ROC) curve and area under curve (AUC), decision curve and calibration curve were applied respectively to evaluate the discrimination, accuracy and application value of the model. Results Six variables, i.e., MA, GLU, CHOL, smoke, history of acute ischemic and systolic pressure were selected to establish the AIS diagnostic model and draw the nomogram. The AUC of the diagnostic model was 0.909（95%CI:0.860 5, 0.945 3）.Calibration curve showed its good reliability of forecast and goodness of fit（Homer-Lemeshow, P=0.104）. Decision curve confirmed its pronounced application value when the probability threshold ranged between 0.6 and 1.0. Conclusion The MA of TEG and such clinical indexes as GLU, CHOL, and history of acute ischemic and systolic pressure could establish a diagnostic model of AIS, providing laboratory basis for the prevention and early diagnosis of AIS.

Key words: Thromboelastogram（TEG）, Acute, ischemic, stroke（AIS）, Logistic, regression