ABSTRACT
@# Objective To compare performance of C5.0 decision tree models and radial basis function(RBF) neural network in predicting the risk of hemorrhagic transformation in acute ischemic stroke. Methods Patients with acute ischemic stroke admitted to hospital were enrolled. Hemorrhagic transformation group and non-hemorrhagic transformation group were divided according to whether hemorrhagic transformation occurred within 2 weeks after admission. Retrospectively collected patients’ case information. C5.0 decision tree models and RBF neural network model were established with the ratio of 7 :3 for training set and test set, and the prediction performance of the model was compared. Results A total of 460 patients’ case information were collected and divided in 314 training set samples and 146 test set samples. Accuracy rates of the C5.0 decision tree model were 96.5% and 80.1%, sensitivities were 98.1% and 82.6%, specificities were 94.8% and 77.9%, Kappa index were 0.93 and 0.60, and AUC were 0.97 and 0.80. Accuracy rates of the neural network model were 72.6% and 74.7%, sensitivities were 87.6% and 88.4%, specificities were 56.9% and 62.3%, Kappa index were 0.45 and 0.50, and AUCs were 0.72 and 0.75. In the training set, the prediction performance of the C5.0 decision tree model was superior to the RBF neural network model. However, there was no statistical difference in the test set.Conclusion C5.0 decision tree model is better than RBF neural network model in risk prediction.
ABSTRACT
Objective To investigate the risk factors and establish the Cox' s regression model on the recurrence of ischemic stroke. Methods We retrospectively reviewed consecutive patients with ischemic stroke admitted to the Neurology Department of the Hebei United University Affiliated Hospital between January 1,2008 and December 31,2009. Cases had been followed since the onset of ischemic stroke. The follow-up program was finished in June 30, 2010. Kaplan-Meier methods were used to describe the recurrence rate. Monovariant and multivariate Cox' s proportional hazard regression model were used to analyze the risk factors associated to the episodes of recurrence.And then, a recurrence model was set up. Results During the period of follow-up program, 79 cases were relapsed,with the recurrence rates as 12.75% in one year and 18.87% in two years. Monovariant and multivariate Cox' s proportional hazard regression model showed that the independent risk factors that were associated with the recurrence appeared to be age (X1)(RR=1.025,95% CI: 1.003-1.048),history of hypertension (X2) (RR= 1.976, 95% CI: 1.014-3.851), history of family strokes (X3) (RR=2.647,95%CI: 1.175-5.961), total cholesterol amount (X4) (RR= 1.485,95%CI: 1.214-1.817), ESRS total scores (X5) (RR= 1.327,95%CI: 1.057-1.666) and progression of the disease (X6) (RR= 1.889,95%CI: 1.123-3.178). Personal prognosis index (PI) of the recurrence model was as follows: PI=0.025X1 + 0.681X2+ 0.973X3 + 0.395X4+ 0.283X5 + 0.636X6. The smaller the personal prognosis index was, the lower the recurrence risk appeared, while the bigger the personal prognosis index was, the higher the recurrence risk appeared. Conclusion Age, history of hypertension, total cholesterol amount, total scores of ESRS, together with the disease progression were the independent risk factors associated with the recurrence episodes of ischemic stroke. Both recurrence model and the personal prognosis index equation were successful constructed.